Gluten Development Explained: Kneading, Autolyse and Bread Texture

Gluten development is the process that turns a shaggy mix of flour and water into a dough that can stretch, hold gas and bake into an airy loaf. It depends on water, mechanical work and time, and on what else is in the bowl: salt, bran and yeast all change how the network forms. This guide explains the science in plain terms, looks at what research says about kneading and resting, and shows how bakers use an autolyse and the windowpane test to read their dough.

What gluten is

Gluten is not something already sitting in flour as a ready-made network. It is a structure that forms when certain wheat proteins meet water. A 2022 review in Foods notes that gluten proteins make up about 80% of the protein in wheat flour, and that they fall into two groups that do different jobs (Zang et al., Foods, 2022):

• Glutenins link together into long chains held by disulfide bonds (bridges between sulphur atoms in the amino acid cysteine). They form the backbone of the network and give dough cohesion and elasticity, the springiness that pulls it back when you stretch it.
• Gliadins are smaller, single molecules that attach to the glutenin network mainly through weaker hydrogen bonds. Once hydrated they behave like a viscous liquid and act as a plasticiser, giving dough its extensibility, the ability to stretch without tearing.

The balance between the two matters. The same review links the gliadin-to-glutenin ratio to dough viscosity. A good bread dough needs enough elasticity to trap gas and spring back, and enough extensibility for gas bubbles to expand and for the baker to shape it.

Step one: water

Without water, gluten proteins cannot move or bond. As flour hydrates, the proteins can begin to interact and bond. Other flour components compete for that water. The Foods review notes that damaged starch absorbs water more readily than intact starch and can take water that gluten would otherwise use, which can disrupt network development.

This is one reason hydration levels in recipes vary between flours, and why water quantities from one recipe do not always transfer to a different flour. If you are new to baker's percentages, our guide to sourdough bread chemistry and crumb structure explains how to calculate hydration.

Step two: kneading and mixing

Kneading does two things. It brings water and flour into contact quickly, and it stretches and aligns the proteins so they form a continuous network. What happens at the molecular level is more interesting than "building" gluten.

A 2021 study in Molecules tracked the glutenin macropolymer, the large linked fraction of glutenin that underpins dough strength, through different mixing and resting times (Feng et al., Molecules, 2021). The authors found that:

• Mixing broke the macropolymer into smaller pieces. Increasing mixing from 3 to 5 minutes cut its content by about 20%.
• Resting after mixing partly rebuilt it. The macropolymer content and the number of disulfide bonds rose with rest, peaking at 60 minutes in their test conditions.
• Resting did not fully reverse the effect of mixing, and a longer rest of 90 minutes reduced both measures again.

The authors concluded that mixing rearranges the protein structure and resting partly restores its chemical bonds. That fits what bakers see: dough kneaded and then left to rest becomes smoother and stronger without further work. The exact times in that study apply to its flour and lab mixer, not to your kitchen, but the pattern of "work, then rest" is useful.

Hand kneading, folds and no-knead methods

Because time and rest contribute to gluten structure, bakers have several ways to reach a well-developed dough:

Autolyse: what it is and what it does

An autolyse is a rest given to flour and water alone before the other ingredients go in. King Arthur Baking's explainer credits the method to Professor Raymond Calvel, a French bread scientist, who developed it in 1974 in response to what he saw as declining French bread quality after more aggressive mixing became common (King Arthur Baking, Using the autolyse method, 2017).

How to do it

1. Mix the flour and water just until no dry flour remains. Liquid sourdough starters and pre-ferments can usually go in at this stage, since they bring their own water.
2. Leave out the salt, the yeast and any stiff starter. Salt and yeast can tighten the dough or be hard to mix in later; the article suggests sprinkling salt on top so you do not forget it.
3. Cover and rest, typically for 20 to 60 minutes.
4. Add the remaining ingredients and continue mixing or kneading.

Why it helps

The same article describes three effects. The flour hydrates fully, and bran softens so it interferes less with gluten, which especially helps whole grain doughs. Protease enzymes in the flour break down some protein, which makes the dough more extensible. And amylase enzymes start converting starch into sugars that yeast can use. Because bonds start forming during the rest, the dough then needs less kneading. The extra extensibility is linked to easier shaping and a more open crumb.

The Molecules study above offers a partial explanation for the "rest makes dough stronger without work" effect, even though it measured rest after mixing rather than a classic autolyse. It is fair to say the evidence for autolyse comes mainly from baking practice; controlled studies of the technique itself are limited.

When to skip it

• Rye-heavy breads: the King Arthur article advises against it for sourdough rye, because rye ferments faster and does not develop gluten in the same way.
• Instant yeast doughs: the author found instant yeast did not dissolve well in fully hydrated dough; adding it during the autolyse or in reserved water are workarounds.
• When you are short of time: a 30-minute rest after mixing everything still brings some of the benefit.

Salt: it slows gluten at first, then strengthens it

Salt is not just for flavour. A 2019 paper on dough rheology in Chemical Engineering Transactions summarises the research: salt increases a dough's resistance to mixing, improves extensibility, reduces stickiness and helps stabilise fermentation, while also delaying gluten network formation because it slows how quickly gluten hydrates (Fanari et al., Chemical Engineering Transactions, 2019). The same paper cites a study in which mixing time rose from 2.75 minutes without salt to 4.75 minutes with 2.4% salt (by flour weight), and notes that results on salt's effect on dough elasticity are not entirely consistent between studies.

In practice, this explains two common habits: adding salt after an autolyse so the flour hydrates first, and expecting a salted dough to take a little longer to come together.

Whole wheat and bran

Whole wheat bread is usually denser than white bread. A common explanation is that sharp bran particles cut gluten strands. Research from KU Leuven suggests a more complicated picture. In a study using model doughs, De Bondt and colleagues found that a proper gluten network could still form when bran was present, but bran still reduced the dough's "strain hardening", a property that helps gas cells grow evenly and resist merging, even after water and mixing time were optimised (De Bondt et al., KU Leuven). Bran that held more water and had a larger surface area had a stronger negative effect. The authors point out that earlier work had already shown lower loaf volume with bran is due to poorer gas retention, not reduced gas production.

For home bakers, the practical points are:

• Adjust water for whole wheat doughs rather than copying a white-flour recipe; the Leuven work shows water level and mixing time both affect the result.
• An autolyse or a long rest gives bran time to hydrate.
• A partly whole wheat loaf will usually rise more than a 100% whole wheat loaf.

Chapati, paratha and bread need different doughs

In the Indian subcontinent, wheat is mostly eaten as unleavened flatbread made from atta (whole wheat meal), as the study below notes. A study by researchers at Punjab Agricultural University notes that the grain quality needed for a flatbread such as chapati differs from that needed for pan bread, and cites earlier work concluding that a medium-strong dough is ideal for chapati (Kumar et al., Journal of Wheat Research, 2013). The study found that dough-mixing measures that correlate well with bread quality were of little importance for chapati quality in the varieties tested.

That matches what cooks know by feel: a roti dough needs to be soft and pliable, and is usually rested for a while before rolling, rather than kneaded to a high-gluten windowpane. Trying to make a tall sandwich loaf from chakki atta alone is a different challenge from making soft rotis with it.

Reading your dough: the windowpane test

The windowpane test is the standard way to check gluten development by hand. King Arthur Baking describes it like this (King Arthur Baking, What is the windowpane test, 2022):

1. Pinch off a piece of dough about the size of a ping-pong ball.
2. Flatten it between your fingers and let it rest for a couple of minutes.
3. Gently pull it out from opposite sides, turning as you go.
4. Hold it to the light. Well-developed dough stretches thin enough for light to show through before it tears.

The article stresses that this is not a pass-or-fail test. It shows how far along the dough is, and the right target depends on what comes next:

Gluten and the finished loaf

Once the dough goes into the oven, heat fixes the structure that the gluten network and trapped gas have built. The crust browns through the Maillard reaction and caramelisation, explained in our guide to the Maillard reaction. The texture you get, open or tight, chewy or soft, is decided long before that, by how you hydrated, mixed, rested and fermented the dough.

The bottom line

Gluten development is the joint work of water, mechanical action and time. Glutenins provide strength and gliadins provide stretch. Kneading rearranges the network and resting lets bonds re-form, which is why short kneads with rests, stretch and folds and an autolyse all work. Add salt after the flour has hydrated, give whole wheat doughs more water and time, and use the windowpane test as a guide to how far along your dough is, not as a pass mark.

Frequently asked questions

How long should I autolyse?

Most guidance suggests 20 to 60 minutes of flour and water before adding salt and yeast. Longer rests are possible, but the benefit depends on the flour and recipe.

Can you over-knead dough?

Research shows longer mixing breaks the glutenin network into smaller pieces, and resting only partly restores it. Watch dough closely when using a powerful machine, and stop once it reaches the development you need.

Why does salt make my dough tighter?

Salt slows gluten hydration at first, then generally increases the dough's resistance and strength. Adding it after an autolyse gives the flour a head start on hydration.

Why is my whole wheat bread dense?

Bran changes dough behaviour and lowers gas retention, even when a gluten network forms. Adjusting the water, resting the dough before kneading so bran can hydrate, and blending in some refined flour are the usual levers.

Does sourdough or autolyse remove gluten?

No. Wheat bread made with any of these methods still contains gluten and is not suitable for people with coeliac disease.

Sources

1. Recent Advances in the Study of Wheat Protein and Other Food Components Affecting the Gluten Network and the Properties of Noodles, Zang P. et al., Foods 11(23):3824, 2022.
2. Dynamic Changes in Glutenin Macropolymer during Different Dough Mixing and Resting Processes, Feng Y. et al., Molecules 26(3):541, 2021.
3. Using the autolyse method, King Arthur Baking, 2017.
4. What is the windowpane test for bread dough?, King Arthur Baking, 2022.
5. The Rheological Properties of Semolina Doughs: Influence of the Relative Amount of Ingredients, Fanari F. et al., Chemical Engineering Transactions 76:703–708, 2019.
6. Selective modification of wheat bran affects its impact on gluten-starch dough rheology, microstructure and bread volume, De Bondt Y. et al., KU Leuven (author manuscript), accessed 2026.
7. Evaluation of wheat genotypes for farinographic characteristics in relation to their chapati-making properties, Kumar S. et al., Journal of Wheat Research 5(1):27–34, 2013.
8. Definition and Facts for Celiac Disease, US National Institute of Diabetes and Digestive and Kidney Diseases, accessed 2026.