Why It Works
- High-protein bread flour produces noodles with plenty of chew and enough elasticity in the dough to prevent breakage when stretching.
- Nutritional yeast contains glutathione, a naturally occurring dough reducer that relaxes gluten and produces a dough with repeatable extensibility in a fraction of the time that it takes using a more traditional method.
- Oil keeps the dough from sticking to your hands when pulling noodles, and also contributes to its overall extensibility.
- Repeated twisting and stretching aligns gluten in a linear direction, so that each successive stretch becomes straighter, more even, and increasingly seamless.
If you’re a fan of Hong Kong cinema, you might be familiar with the 1997 action comedy Mr. Nice Guy, starring Jackie Chan. In the film’s opening scene, Jackie is on live television, standing at a flour-dusted table, stretching, twisting, and pulling a piece of dough into fine strands of noodles, a process the TV host can only describe as alchemy. “When I first saw [that movie],” recalls Luke Rymarz, a software engineer and hand-pulled noodle enthusiast based in San Jose, “I thought, ‘Oh cool, special effects. That’s neat.’ And then I realized that hand-pulled noodles were actually a thing. And I thought, all right, time to figure out how to do this.”
Luke spent the next year experimenting and documenting those experiments, blindly tinkering his way toward a passable recipe for lamian. Lamian translates to “pulled noodles,” and specifically refers to the thin variety famous in China. “At the time—this was 12, 13 years ago—there was hardly anything on the internet in English, or any information about it. All of my experience is [based on] the little bits I could glean off of Google translate. And just making dough, over and over again, for hours.”
Like Luke, I have spent the last few years learning how to make lamian. I first encountered these noodles at a tiny hole-in-the-wall lunch joint in Montreal. A rail-thin cook was throwing, twisting, pulling, and generally manhandling a mound of dough nearly the length of his body. Within seconds—after a few nimble flicks and plenty of flair—he created hundreds of strands of noodles. He tossed the noodles in a boiling cauldron of water, reached for another log of dough, and repeated this dance for the remainder of my meal—no hesitation, no breaks, not even a paltry sip of water. Talk about a real dough-slinger.
Of course, this guy made pulling those noodles look easy, even natural. It’s neither of those things. In fact, I’m willing to bet that anyone who has ever tried to make thin pulled lamian at home, with no prior experience, will tell you: It's hard. Recipes and videos exist online, but there’s a swamp of conflicting information, most of which leads down a treacherous black hole teeming with Reddit neckbeards, disgruntled YouTube commenters, and holier-than-thou Chowhound geeks.
Pulling noodles is something I never imagined pursuing. The technique seemed unattainable, unapproachable, reserved for masters of a time-honored and mystical craft—certainly inappropriate for home cooks. But as luck would have it, while working at my last test kitchen job with Sasha years ago, I was assigned to learn about pulled noodles in all their forms. After slurping virtually every pulled noodle Boston had to offer, I set about developing recipes. Lamian sat squarely among them. But after almost 50 tests, I still didn’t have a working recipe. Even building on the work of people like Luke, my method was fickle, riddled with inconsistencies, and difficult to replicate. The recipe never saw the light of day.
Time passed. I tried, here and there, to tinker with my formula. I watched videos. I spent hours reading archived forum posts from the furthest reaches of the web. I reached out to family members to help translate texts. I tried interning at noodle shops (the owners of which either shunned me or swore me to lifelong secrecy). I reached out to Luke for guidance. I was determined to figure out a recipe that would work for the home cook—no special equipment, no advanced noodle degree, no secret ingredients, no secrecy required.
Then I had a breakthrough. And it all came down to understanding gluten.
How Hand-Pulled Noodles Are Typically Made
The de facto Mecca of pulled noodles is Lanzhou, the capital city of China’s Gansu Province. The city is the birthplace of Lanzhou beef noodle soup, a dish that’s inspired the creation of over 35,000 noodle shops across China. You can even attend one of a handful of noodle-pulling schools, with programs ranging in duration from a few days to a few years. There, students learn the ins and outs of making lamian, always by hand, starting from mixing and kneading dough all the way to pulling and cooking. In order to be certified, students must pass a final exam: Roughly 20 minutes to produce nine varieties of noodles of different sizes and shapes.
The process is simple: Wheat flour, water, salt, and sometimes an additive are mixed and kneaded until a pliable dough forms. This kneading stage takes anywhere from 15 minutes to over an hour, depending on the formula and the practitioner. The dough is stretched, doubled over, twisted, and stretched again, and this process is repeated until the dough can be stretched easily to arm’s length. The dough is then rolled in flour or oil and pulled into thin strands. The key to success is repeatable extensibility—the ability to stretch the dough over and over without it breaking or tearing—which, as we’ll see, defies many of the commonly understood laws of gluten.
Sometimes additives like fat, alkalis, or other ingredients are incorporated into the dough to improve texture or to facilitate extensibility. The most common and traditional additive is a mysterious powder called penghui, which is derived from a desert plant called penghuicao (halogeton). The plant is roasted and processed with potassium carbonate and packaged as a commercial powder. Minute quantities of penghui are mixed with flour, resulting in a dough with ideal extensibility and chew. One source claims that the addition of penghui transforms the dough to putty, allowing the cook to stretch the dough seemingly at will.
The problem? Unless you or someone you know lives in China, finding penghui isn’t easy. And seeing as I didn’t feel comfortable asking people to import a sketchy white powder across international borders, I was left to find other ways to develop that all-important repeatable extensibility.
Gluten’s Role in Dough: Beyond the Basics
To fully understand my noodle journey, it’s useful to know a bit about dough, and specifically gluten. If you’ve ever dealt with a dough in your life, you probably know something about gluten. Simply put, gluten is the protein matrix (basically a molecular web) that forms when wheat flour is mixed with water. Gluten is what gives wheat flour dough its structure and properties: The stretch, the texture, the snap, the stickiness, and flow.
If you’ve delved a little deeper, you might also know that gluten comprises two protein molecule units: glutenin and gliadin. The larger glutenin molecules are mainly responsible for the strength and elasticity of a dough. Look at them closely enough and you'll see they are kinked, linear, and chain-like, and have many sites for bonding with other glutenin molecules.
Smaller gliadin molecules, on the other hand, are said to contribute to the extensibility of a dough. They are spherical (or “globular”), have less surface area, and have limited capacity for bonding with other molecules.
Together, these two proteins are what give dough its viscoelastic properties—meaning it can simultaneously flow (viscosity) and be elastic.
Before going any further, let’s also define elasticity and extensibility—both of which are important to understanding noodle-pulling.
Elasticity
Elasticity refers to a dough’s ability to return to its initial position after deformation. If you stretch a dough and it snaps back to its original shape, that dough is said to be elastic. Elasticity is important because it gives dough both structure and integrity. Without elasticity, a bread dough wouldn’t hold its shape or have any chew. Similarly, without elasticity, a noodle dough would fall apart if stretched too far, and likely disintegrate when cooked. On the other hand, a dough with too much elasticity would tear under too much stress, like a rubber band snapping.
Elasticity is the result of those large, chain-like glutenin molecules bonding together. Mixing, kneading, and resting dough all encourage this bonding. The more bonds between glutenin molecules, the more elastic a dough becomes.
(In some circles, dough nerds like to distinguish tenacity from elasticity. Tenacity is the property of a dough to resist a stretching action or deformation in the first place. You might encounter this phenomenon while trying to shape or stretch a low-hydration bread dough, for instance. For our purposes, this distinction is not too important. But the more you know, you know?)
Extensibility
Extensibility is defined as the ability of a dough to stretch. For a pulled noodle dough like lamian, extensibility is crucial. In general, the more compact and spherical gliadin protein molecules contribute to extensibility because they move freely within the gluten matrix. They do not bond in the way that the larger glutenins do.
The balance between elasticity and extensibility determines how well a noodle dough fares. In general, the ideal noodle dough has high extensibility and just enough elasticity to retain its structure and chew.
Bonds Between Gluten Molecules
The last and most critical concept—the one that most people gloss over—has to do with bonding between gluten molecules. Both glutenin and gliadin involve disulfide bonds. For gliadins, these bonds occur within single molecules to stabilize their spherical structure. But for glutenins, disulfide bonds (a.k.a. disulfide bridges) can occur between their larger, chain-like units. These bonds are strong, and they’re believed to determine the elasticity of a dough. In other words, the more disulfide bonds there are between glutenin units, the harder a dough is to stretch.
Think of a single glutenin molecule as a train with several train cars hooked together. Now imagine a second train on a parallel track with chains joining it to the first train. Those chains are disulfide bonds. That’s basically what occurs between glutenin units, but in multiple directions and multiple orientations, forming a complex, strong web that is difficult to break.
Luckily, these bonds can be broken.
Testing
Trial 1: Hand Mixing, No Additives
By certain accounts, it was possible to pull noodles through sheer force of will and a little elbow grease. Apparently, you could even do it without additives. This was the first strategy I tested years ago: I mixed flour, water, and salt, and kneaded until the dough came together. I kneaded, tore, and stretched the dough further until I could get some semblance of extensibility. On average, this process took anywhere from 40 minutes to over an hour of sweaty, rage-inducing labor. At best, I could get the dough to stretch a couple times before tearing. At worst, I was stuck in dough purgatory, cursed with a dough that refused to relax and tore at the slightest suggestion.
I tinkered with hydration, flour type, and salt concentration. I tried no less than 35 times. But no matter which formula I used, I couldn’t readily produce a dough with repeatable extensibility. I did have some success with low-protein cake flour, which seemed to relax after persistent kneading and twirling, but I was still nowhere near pulling passable noodles.
One source recommended an overnight rest after initial mixing. In theory, this extended rest allows the dough to hydrate and gluten to develop fully, and relaxes gluten enough to facilitate stretching (a familiar phenomenon if you’ve ever made fresh egg dough pasta). There is also some enzymatic breaking of disulfide bonds, which contributes to extensibility. At first, this technique seemed promising. I could stretch the dough a few times easily. But after a few more passes, the dough eventually snapped back and tore in my hands. I still wasn’t producing enough extensibility.
If hand-mixing didn’t work—or wasn’t feasible in a sensible amount of time—then how else could I get an extensible dough?
Trial 2: Really, Really Aggressive Kneading
The second strategy I found involved kneading a dough in a stand mixer for an extended period of time. The logic here was that aggressive, sustained kneading weakened the gluten network sufficiently to produce a plastic dough with low elasticity and high extensibility. (Initial mixing and kneading develops gluten to a point, producing a decidedly un-stretchy, elastic dough. But excessive kneading beyond that point could actually weaken the gluten network.) More precisely, those strong disulfide bonds could be broken mechanically; with sufficient force over time, enough bonds could be broken to produce a dough with repeatable extensibility. And instead of the back-breaking work of stretching and kneading by hand, a stand mixer could accomplish this more efficiently and in less time.
Anecdotally, this logic made sense to me. I remembered spotting a giant Hobart mixer at that noodle shop in Montreal, whirring away, kneading dough for the entirety of my meal. I even found interviews of noodle chefs in which they admitted to using large bread mixers to meet the demands of a large operation. In the most extreme example, I found a recipe that recommended kneading a dough for nearly two hours in a stand mixer.
And you know what? This method definitely worked. After two hours of kneading in a KitchenAid, I found this “over-kneaded” dough to be pliable, extensible, and similar to silly putty in texture. There was very little elasticity. Just a few minutes of stretching and twirling later, I was able to pull some respectable noodles.
But I still wasn’t satisfied. After all, I couldn’t justify the risk of burning out the motor of a pricey stand mixer just to make some noodles. I didn’t want to leave legions of would-be noodle pullers with $300 paper weights. Plus, two hours is a criminally long amount of time. Ain’t nobody got time for that.
Fortunately, there exist recipes with more moderate kneading times. The most popular of these comes from Luke Rymarz himself. The key feature is a 12-minute kneading time in a stand mixer, followed by another 15 minutes of stretching and twirling by hand. “I was doing it all by hand initially,” he says. But one day, after posting his recipe online, he was invited to demo his technique at a nearby cooking school. The instructor prepared his dough in a huge industrial mixer. “When he gave me that dough, I went, ‘Holy cow. This is amazing. This is the best dough.’ It was warm, it had a nice texture, it was like clay. Very pliable.”
Up until that point, Luke’s recipe was the closest I had come to successfully and consistently pulling noodles within a reasonable amount of time. The resulting dough was pliable, extensible, and fairly easy to pull into thin noodles. But the technique still had some disadvantages: The cooked noodles lacked the chew and structure I was after; the recipe required very specific brands of flour; and the technique demanded extensive stretching and twirling for success, despite all that time kneading in a stand mixer. The recipe had too many variables, and took too long to make. It was inconsistent. Most of all, I wanted a simpler method that didn’t rely so heavily on machinery. “I’ve repaired my stand mixer probably four or five times,” Luke warns. “So it’s not a great option for the home cook.”
Trial 3: Alkali Madness
Since I couldn’t rely on machinery, I turned toward uncharted territory: additives. Using penghui was out of the question—I just couldn’t get my hands on any. But could there be any substitutes?
Popular opinions online suggested that it was the alkali in penghui that changed the texture of the dough and made it extensible. In reality, adding an alkali to a dough just made it tougher and harder to stretch. According to Luke, who tried every possible combination of kansui (a popular alkali solution used in ramen noodles), baking soda, and baked baking soda (sodium carbonate), “It helps with texture, and gives you a better chew, but it makes it harder to pull.”
I found his observation to be largely true. I tried adding baking soda and lye water in varying concentrations to a working dough recipe, kneading by hand and stretching until I could stretch the dough. As the concentration of alkali increased, the dough became increasingly difficult to stretch—in many cases tearing under stress.
If adding alkali wasn’t relaxing a noodle dough, then what was it doing, and how? A study in the Journal of Cereal suggests that the addition of kansui increases disulfide bond formation between glutenin units. More disulfide bonds mean more elasticity, and more elasticity means more chew at the expense of extensibility. Long story short, it appears that adding alkali makes pulling noodles harder, not easier.
Trial 4: The Wide World of Dough Reducers
Adding an alkali wasn’t the answer. I had to dig deeper. Based on one translated source, the composition of penghui includes sodium, potassium, and sulfur. And given my previous tests with potassium carbonate and sodium bicarbonate (both alkalis), I was willing to rule out the first two ingredients.
But sulfur? A quick search for sulfur additives in dough led me to a comprehensive post describing the role of dough reducers in industrial breadmaking. Dough reducers (or reducing agents) are a class of dough conditioners used to decrease mixing time and to improve extensibility. Chief among them are the protein-based reagents cysteine and glutathione—and, coincidentally, sulfites.
I reached out to the author, Dr. Jacinthe Côté, a food biotechnologist and Product Management Director at Lallemand Inc., a major research, development, and manufacturing company that specializes in yeasts and bacteria for natural fermentation processes, including industrial dough processing. According to her, reducing agents act like mixing to reversibly break down the gluten network, resulting in increased extensibility. “You have to be aware that those [disulfide] links can be broken. If you manipulate the dough, or if you stretch it too much, the bridge can be broken. It can be done mechanically, or it can be done chemically.”
For Dr. Côté, gluten development during mixing is best thought of as an oxidation-reduction reaction, hence the name “dough reducers.” “The formation of a disulfide bridge is an oxidation reaction...involving gluten [proteins] containing sulfhydryl (or thiol, denoted by -SH) groups. The oxidation process links the sulfhydryl groups on one protein with another.” Breaking those disulfide bonds chemically is a reduction process known as disulfide interchange. Do you need to know all that hard science? Not really. What’s important is this: Each reducing agent participates in some form of disulfide interchange to relax a dough. But they aren’t all created equal.
Sulfites
Sulfites are a common reducing agent used in cookie, cracker, and biscuit production; they are used to relax doughs for shaping. These reagents act like caps, covering the reactive sulfhydryl groups on gluten proteins. This capping makes it difficult to reform disulfide bonds—resulting in a dough that can’t form a strong gluten network in exchange for extensibility.
The most commonly available form is a salt, such as sodium metabisulfite (which you can find at most homebrew stores as a preservative for wine making). Unfortunately, sulfite salts have the distinct odor of rotten eggs, can cause sensitivity reactions in some people (rashes, hives, indigestion, as some sulfite-sensitive wine drinkers may know), and are closely regulated by the FDA. They are effective at levels between 20-100 parts per million, so even if you used 1/16th of a teaspoon of a sulfite salt, you would need to add it to several pounds of dough to be effective. Ultimately, sulfite salts are not suited for home cooking (unless you’re making a huge batch of dough).
I had read that penghui smelled strongly of rotten eggs. Based on its reported composition and smell, I suspect that penghui contains some kind of sulfite salt, which would explain its dough-relaxing properties. In the end, I didn’t think sulfites were a viable—or healthy—option for the home cook.
Cysteine
Cysteine is an amino acid, and the go-to reducing agent in commercial bread production. It’s cheap to produce, you can add it directly to a dough, and it acts quickly, reducing the number of disulfide cross-links between glutenin chains via disulfide interchange. Like sulfite salts, cysteine must be used in small amounts (10-90 parts per million), so it’s not the most practical choice for home cooks looking to make a single loaf of bread or a small batch of noodles. While you probably won’t find it in the baking aisle at your local grocery store, you can buy cysteine over the counter as a dietary supplement. But it does have a bad rap among health-conscious consumers: Most cysteine is extracted from the feathers of birds and hog hair. “There’s a big chemical extraction process,” explains Dr. Côté, and it results in a product that hardly resembles feathers or hair at all.
Despite the bad reputation, I decided to give cysteine a shot. I split open a pill of L-cysteine and sprinkled 0.1 grams of the powder into a working recipe of bread flour, salt, and water. The smell of metallic, faintly rotten eggs immediately stung my nostrils. Within minutes of kneading, the dough turned to a puddle in my hands. I could stretch the dough seemingly infinitely. But the dough lacked any semblance of structure, making it impossible to pull noodles that could hold their shape. I had added too much cysteine. Clearly, using cysteine could work to relax my dough, but it was also impractical and difficult to use correctly: I couldn’t justify asking home cooks to buy a whole bottle of cysteine just to use a couple milligrams of the stuff.
Glutathione (a.k.a. The Winner)
Glutathione is a peptide (a peptide is smaller than a protein, usually containing less than 50 amino acids) that contains cysteine, and functions in a similar way to relax doughs. Most commercial glutathione comes from natural sources like heat-treated, inactive yeast. “When you’re drying the yeast, you’re stressing [it] a little bit, and some of the cells die,” says Dr. Côté. “When the yeast cells die, they release some of their components that are inside the cell. One of the components that is naturally occurring is glutathione.” Lallemand specializes in a product called Fermaid, a non-leavening yeast product that is abundant in glutathione. Still, I didn’t want to buy a specialty, industrial-grade ingredient just to make some noodles (like sodium metabisulfite, you can sometimes find Fermaid at homebrew stores).
But what about nutritional yeast? Technically, it’s deactivated yeast, too, so it would have some proportion of glutathione. Could it work to relax a noodle dough? I pitched the idea to Dr. Côté. She lit up. “Ok! That’s a good one, too. That logic makes sense. A lot of bakeries just use regular inactive yeast (which is generally available industrially). I think there’s about 1.5–2% glutathione in that product, and it works.” You can make inactive yeast by slowly cooking or drying out fresh yeast, but the process is time-consuming. And if nutritional yeast was readily available and just as good, it seemed like a better option.
I started testing with a tablespoon of nutritional yeast mixed into my dough. Unlike my tests with cysteine, I found I could use reasonable amounts of nooch without worrying about over-relaxing the dough. To speed mixing up, I combined all the ingredients in a food processor and ran it just until a dough formed, then transferred the dough to the counter. After a few minutes of kneading and twisting, the dough relaxed considerably and became more extensible.
But unlike the dough with cysteine, it still retained its structure. As I dialed up the amount of nutritional yeast, the dough became even more extensible. By the fifth trial, I was able to pull noodles. The cooked noodles were ideal: chewy, even in thickness, and slurp-worthy. Using nutritional yeast also had a couple unintended benefits: It tinted the noodles a pale yellow—reminiscent of Japanese alkali ramen noodles—and it imparted subtle umami flavor (nutritional yeast is rich in glutamate, giving it a savory, cheesy quality, which explains why it’s often used as a stand-in for Parmesan in vegan recipes). And the best part? I went from mixing to pulling and cooking noodles in 15 minutes flat.
Finally, I had found a reliable way to pull noodles—or at least a way to get a dough with repeatable extensibility. And I didn’t need special equipment, special ingredients, or years of noodle school training.
Refining the Formula
I solved the hardest part of the noodle problem: I had extensible, cooperative dough that could be both pulled repeatedly and made very quickly. All that remained was to hone in on a foolproof formula for optimal handling and texture. I’ll break the rest of my testing down briefly, ingredient by ingredient:
Flour
Flour is by far the most important ingredient in any dough. I tested various brands of low-protein cake flour, all-purpose flour, and bread flour. Each of these flour types vary in their gluten potential (gluten potential is a term for how much gluten development is possible). Cake flour sits at 7-9% protein content (protein content includes proteins like albumin and globulin, in addition to gluten proteins); at the other end of the spectrum, bread flour contains 12-14% protein, depending on the brand. A flour with higher protein content tends to form a dough with more gluten potential and, consequently, more elasticity and chew.
In the end, I found that bread flour resulted in noodles that pulled easily but had the proper elasticity for structure. (I used King Arthur bread flour, which has the highest protein content available of all common flours you can find at grocery stores.) The noodles held their shape well during pulling, and separated into even strands. The cooked noodles were chewier and had a more pleasant spring compared to noodles made with other flours. As the protein content of the flour decreased, dough handling and chew worsened. For instance, the same formula with cake flour was stickier and tended to droop and stick to my hands while pulling and stretching; the resulting noodles were uneven, lacked chew, and were spongy. I also observed that higher-protein flours were able to absorb more water without becoming sticky and unmanageable.
Water
In general, water serves two primary functions in a noodle dough: It is essential to hydrating flour to form a stable gluten network; and it facilitates extensibility, making a dough stretchable. It’s useful to think of dough as a suspension of solid particles (starches) in a viscous fluid. If you add more water to that dough, you are increasing the size of the suspension, giving more space for the solid particles to move around, which means the dough will stretch more. At the same time, adding more water makes a dough stickier, harder to handle, and less elastic. I found the ideal hydration for my noodle dough (as a percentage of flour weight) sat between 62 and 68%. At levels lower than this range, the dough tended to be too elastic and too resistant to stretching. At a hydration higher than 68%, the dough was more extensible, but stuck to the work surface and to my hands, and was difficult—if not impossible—to stretch evenly.
Salt
Salt is used mostly for flavor in my formula. But. depending on the concentration, salt also acts as a conditioner in a dough, strengthening the gluten network. I tested doughs with and without salt. I found that without salt, doughs tended to be more slack and sticky. Adding salt made doughs more elastic, but easier to handle.
Oil
Oil has several effects on dough handling and cooked noodle texture. In general, adding oil inhibits gluten development, since a portion of flour absorbs that oil during mixing. Oil also affects viscosity in a similar fashion to water: It makes a dough softer and more extensible. Lastly, oil seems to improve dough handling by mitigating stickiness to surfaces and hands. That final quality was integral to clean, even stretching in my recipe. Adding oil to my dough slightly increased extensibility without having to add even more water, which would have made my dough too sticky to handle.
Nutritional Yeast
For optimal extensibility, I found nutritional yeast to be effective between 5 and 8% of the total flour weight in my tests. As the protein content of flour increased (up to King Arthur bread flour), I could dial up the amount of nutritional yeast to the top end of that range without compromising structure. As the protein content decreased (to say, Pillsbury cake flour), less yeast was required, and additional yeast made the dough too sticky and slack.*
*If you’re curious, nutritional yeast contains 2.5 milligrams of glutathione per gram.
A Note on Alkali
What about all that hype around penghui and kansui? For this recipe, I noticed that alkali did make my noodles chewier; but it also made pulling noodles more difficult, and the cooked noodle shape was wavy, curled, and uneven. My noodles had plenty of chew already from high-gluten bread flour, and the added benefit in texture wasn’t worth the regression in dough handling and shaping. In the interest of keeping things simple, I decided to leave alkali out. You could definitely experiment with adding some alkali to your noodle dough to improve chew, but keep in mind that it will negatively impact extensibility.
Rules for the Road
So now we’ve got a dough formula that works. But it’s still up to you to bring it all home and pull noodles. Here are some guidelines and words of advice to keep in mind for successful noodle pulling.
Use a Scale
You need a scale. I’m not going to pretend otherwise. You might think you could get away with your heirloom tablespoons and that cute chipped porcelain measuring cup that you copped from Goodwill, but trust me when I say: It will do you no good. If you want consistency, use a scale (preferably digital).
Stretch and Twirl
After the formula, the most important aspect of this recipe is proper kneading before pulling noodles. All that stretching, twirling, and doubling over of dough might look flashy, but the process serves a function: It aligns gluten in a roughly linear orientation. Most of the time, general kneading (especially in a mixer) mashes gluten proteins in a random, non-linear way. That randomness is great for breads, which must expand in all directions when rising or baking. But noodles are straight, and need to extend in a linear way. Stretching and twirling is basically linear kneading: You’re mechanically making and breaking bonds in gluten, and aligning them in roughly one direction.
Practice Pulling
Once you’ve sufficiently developed and aligned your dough, it’s time to pull noodles. This stage is a perfect time to practice. Ultimately, pulling noodles is a hand skill that takes some repetition and feeling. “My recommendation is don’t jump the gun and try to boil your noodles. Just practice,” Luke recommends. Generally, flouring the table and rolling the dough in flour helps keep the noodle strands separate when pulling. But once you’ve floured the table, you have to pull noodles and cook them, since additional flour would be worked into the dough, throwing off our intentionally crafted formula.
Fortunately, the un-floured dough has that repeatable extensibility, so you can practice the pulling motions repeatedly, without fear of the dough tearing or snapping back. “Just keep trying to make noodles, over and over again—without actually flouring them and throwing them into a pot and expecting to eat anything. A lot of this is really the dexterity and knowing how to hold the two ends of the dough.”
Don’t Be Afraid
If I’ve done my job, this should be a very forgiving dough. So don’t be discouraged if you can’t get the pulling motions quite right initially. If the dough tears (it shouldn’t), just roll it back up and try again. Don’t stress. If the noodles are uneven, try again. If one or two strands break as you’re pulling, don’t freak out. Remember: It’s just dough.
Go Forth and Pull
At this point, the noodle masters of yore are likely rolling in their graves. Traditionalists out there will probably put me on full blast. “Lamian doesn’t have nooch! How dare you? That’s not authentic.”
Tradition was never the point. Exploration and understanding—they are worth far more. I could have just as easily told you to go out and smuggle in some penghui, knead your dough for hours, and make some authentic noodles. But this is a lamian recipe for home cooks. My goal was to develop a noodle dough that anyone could pull; a way to practice noodle pulling without resorting to prohibitive means or herculean efforts.
Just to make sure, I sent my recipe to Luke for him to try out. He emailed me back a couple days later with pictures and even a video of him pulling noodles. It took him ten minutes from mixing to eating—a personal record. “I’m kinda speechless...huge stamp of approval from me!”
No matter the method, making lamian is magic. It’s a perfect demonstration of the alchemy, and chemistry, of cooking. It embodies the excitement, the thrill, and the spirit of making something amazing out of the ordinary. Learning about the process has been equally enriching. So go out and make some lamian dough. Practice, pull, and practice again. Then you'll be ready to cook them up and slide them into a fragrant bowl of lamb soup.
April 2020
Recipe Details
Hand-Pulled Lamian Noodles Recipe
Ingredients
425g (15 ounces; 3 cups) bread flour, plus more for dusting
28g (1 ounce; 1/2 cup) nutritional yeast (see note)
4g (1 teaspoon) Diamond Crystal kosher salt; for table salt, use about half as much by volume or the same weight
285g (10 ounces; 1 cup plus 3 tablespoons) cool water
1 tablespoon (15ml) vegetable oil
Directions
In the bowl of a food processor, pulse the flour, nutritional yeast, and salt together to combine. With processor running, add water and oil and process until dough forms and runs around processor blade, about 30 seconds.
Transfer dough to clean counter. Knead dough until smooth. Roll, knead, and stretch dough, doubling back after each stretch, until dough begins to stretch without tearing, 3 to 5 minutes.
Stretch dough to arm’s-length, then bring ends together to twirl dough into an even, twisted log (roll dough on counter after twisting to even out thicker areas). Repeat stretching and twirling until dough pulls full arm’s-length with no resistance and no tearing, about 10 minutes longer. Roll dough into an even log shape.
If you plan to pull and cook the noodles immediately, dust the counter liberally with flour and bring a large pot of salted water to a rolling boil over high heat. If you would like to practice the noodle-pulling technique, do not flour the counter and proceed with the pulling instructions. Flouring the noodles is a point of no return—once floured, they must be cooked. If the dough is not floured, you can practice the pulling technique over and over without harming the dough. Gently roll dough log in flour, then use a bench scraper to divide dough into 3 equal portions.
Working with one piece at a time, shape into 15-inch log. Holding log at both ends, stretch to about 30 inches in length. (The following instructions are written for right-handed people; invert them if you are left-handed.) Bring right end of dough to left hand, forming loop. Using free right hand, gently tug and pull dough from midpoint of loop in fluid motion to about 30 inches. Move dough through floured surface to keep from sticking.
Bring right end of dough to left hand and gently bunch dough ends together, forming another loop of 4 strands. Using free right hand, gently tug and pull dough from midpoint of loop in fluid motion to about 30 inches. Move dough through floured surface to keep from sticking. Repeat process 2 to 3 more times until noodles are about 1⁄8 inch in diameter.
Use bench scraper or knife to cut and separate noodles from small clump of dough in left hand. Immediately transfer portion of finished noodles to the pot of boiling water, and cook, stirring with tongs or chopsticks to prevent sticking, until noodles are cooked through but still chewy, 30 to 45 seconds. Drain noodles thoroughly using a fine-mesh strainer or noodle basket, shaking off as much excess water as possible, then transfer to a serving bowl with hot broth.
Repeat stretching, pulling, and cooking steps with remaining two portions of dough. A fourth portion of noodles can be pulled from excess dough scraps if desired. Knead, stretch, and twirl excess scraps of dough into even log, roll in flour, and repeat pulling process. Serve cooked noodles immediately.
Special Equipment
Digital scale, food processor, fine-mesh strainer
Notes
For best results, use a digital scale set in grams to measure ingredients.
Be sure to use nutritional yeast well within its expiration date. Expired nutritional yeast will not work with this recipe—the dough will not relax as intended.
This noodle dough is designed to be kneaded, pulled into noodles, and cooked all in one go. Unlike other noodle recipes, the dough and noodles cannot be made in advance and cooked at a later time.
Make-Ahead and Storage
The noodles are best enjoyed immediately. In fact, the recipe calls for cooking each portion of noodles immediately after pulling, as is customary with lamian noodles.
Read More
Nutrition Facts (per serving) | |
---|---|
437 | Calories |
6g | Fat |
80g | Carbs |
16g | Protein |
Nutrition Facts | |
---|---|
Servings: 3 to 4 | |
Amount per serving | |
Calories | 437 |
% Daily Value* | |
Total Fat 6g | 7% |
Saturated Fat 1g | 3% |
Cholesterol 0mg | 0% |
Sodium 397mg | 17% |
Total Carbohydrate 80g | 29% |
Dietary Fiber 4g | 15% |
Total Sugars 0g | |
Protein 16g | |
Vitamin C 0mg | 0% |
Calcium 24mg | 2% |
Iron 1mg | 6% |
Potassium 169mg | 4% |
*The % Daily Value (DV) tells you how much a nutrient in a food serving contributes to a daily diet. 2,000 calories a day is used for general nutrition advice. |