The windowpane test tells you whether your dough's gluten is developed enough to bake. Pinch off a golf-ball-sized piece and stretch it slowly between your fingers: if it thins into a translucent sheet you can see light through without tearing, the gluten is ready for bulk fermentation. If it tears first, keep mixing or folding and test again. It takes five seconds, needs no tools, and is the first checkpoint that everything downstream depends on.
The windowpane test is the single most reliable way to assess whether your bread dough’s gluten network is sufficiently developed. It takes five seconds, requires no tools, and gives you a clear answer: keep mixing, or move on.
Every major bread author references this test. All agree on what it demonstrates and how to perform it. If you bake bread and you do not know this test, you are guessing at one of the most critical stages of the process.
How to Perform the Windowpane Test
- Pinch off a small piece of dough — roughly golf-ball sized.
- Gently stretch it between your fingers and thumbs, pulling outward from the center. Use both hands, with fingers underneath supporting the dough and thumbs on top.
- Pull slowly and steadily — do not yank. You are coaxing the gluten, not forcing it.
- Look for translucency. Well-developed gluten stretches into a thin, almost transparent sheet. You should be able to see the shadow of your fingers through it, or even read text through it in strong light.
The Windowpane Test, Step by Step
Five seconds, no tools
Pinch off a piece
Tear off a golf-ball-sized piece of dough.
Stretch outward
Pull it between fingers and thumbs, from the center out.
Pull slowly
Coax the gluten -- never yank.
Pulling fast tears the dough at any stage of development.
Look for light
A thin, translucent sheet you can see through means the gluten is ready.
Jeffrey Hamelman describes the target: “When it is sheeting fully, you will be able to coax it out into a thin, almost transparent, sheet.”
What the Results Mean
The Dough Stretches Thin and Translucent
Gluten is fully developed. The glutenin and gliadin proteins have hydrated, cross-linked via disulfide bonds, and formed a strong, interconnected network. The dough is ready for bulk fermentation.
At this stage, the dough should feel smooth, cohesive, and slightly elastic — it springs back gently when you poke it but does not resist stretching.
The Dough Tears Before Becoming Translucent
Gluten is under-developed. The protein network has not yet formed enough cross-links to support a thin sheet. The dough needs more mixing or folding time.
If you are kneading by hand, continue for 3-5 more minutes and test again. If using a stand mixer, continue on second speed for 2-3 more minutes. If you are using a fold-only method (like Hamelman’s six-fold technique), perform 2-3 more folds at 10-minute intervals and retest.
The Dough Stretches But Feels Sticky and Fragile
Possibly over-mixed. This is the danger zone. Excessive mixing can break the disulfide bonds that hold the gluten network together, producing a dough that feels slack, sticky, and weak. It stretches because it has no resistance left — not because the gluten is well-developed.
Hamelman’s overoxidation warning applies here: industrial mixers in post-WWII France destroyed bread quality by overoxidizing dough — breaking gluten bonds and destroying carotenoid pigments. The same thing happens in a home kitchen if you mix too long on high speed.
Signs of over-mixing:
- Dough is very sticky and does not hold its shape
- The stretch is limp rather than elastic
- The dough tears into wet, ragged pieces rather than clean holes
- Color appears paler than expected (carotenoid destruction)
Over-mixing is not fixable. The bread will be flat, dense, and lacking in flavor. This is rare with hand mixing (you would exhaust yourself first) but possible with a stand mixer on high speed.
The Science Behind the Test
The windowpane test works because of gluten’s molecular structure.
Glutenin molecules are large polymeric proteins connected by disulfide bonds (S-S bonds between cysteine residues). These provide elasticity — the tendency to spring back. High molecular weight glutenin subunits (HMW-GS) form the backbone of the gluten network.
Gliadin molecules are smaller, monomeric proteins that act as a molecular lubricant within the glutenin network. Emily Buehler describes gliadin as acting like “a molecular lubricant, allowing the glutenin chains to flow past each other.” Gliadin provides extensibility — the ability to stretch without tearing.
When dough is fully mixed, the glutenin backbone is well-connected (providing structure) and the gliadin is evenly distributed (providing flow). The result: a sheet of dough that stretches thin, transmits light, and holds together. This is the windowpane.
When dough is under-mixed, the glutenin chains have not yet formed enough cross-links. The dough tears at weak points because the network has gaps.
Hydrophobic interactions also contribute. Nonpolar amino acid side chains cluster away from water, creating internal cohesion that reinforces the disulfide bond network.
When to Use the Windowpane Test
After Machine Mixing
This is the most common application. After combining ingredients and mixing on second speed in a stand mixer, check the windowpane every 2-3 minutes starting at about minute 5-6. Most bread doughs reach full development between 6-12 minutes of second-speed mixing, depending on the flour and hydration.
Peter Reinhart’s guideline: combine on low speed with paddle for 1 minute, switch to dough hook on medium speed, and check for windowpane every few minutes for 6-12 minutes total.
After Hand Kneading
Hand kneading takes longer to develop gluten because human hands generate less mechanical energy than a machine. Check the windowpane after 10-12 minutes of kneading, then every 3-5 minutes thereafter.
After Folding (Fold-Only Methods)
If you are using a fold-only method — stretch-and-fold (Robertson), pincer method (Forkish), or Hamelman’s six-fold — the windowpane test works but the result looks different.
Fold-developed dough often produces a thicker, less perfectly transparent window than mixer-developed dough. This is normal. You are looking for a cohesive sheet that holds together when stretched gently — it does not need to be paper-thin.
After 3-4 sets of folds over 1.5-2 hours, most doughs will pass a “good enough” windowpane. If the dough still tears immediately with no stretch, add 1-2 more folds.
After Autolyse
Checking the windowpane after autolyse can tell you how much mixing work remains. A good 30-60 minute autolyse in a high-protein flour may already produce a partial windowpane — meaning gluten has started developing through hydration alone. This is Hamelman’s key insight: autolyse can reduce mixing time by up to 50%.
When to Skip the Windowpane Test
Not every dough needs (or can produce) a perfect windowpane.
Whole wheat dough: Bran particles physically puncture gluten strands. A 100% whole wheat dough will never produce a thin, transparent window — the bran tears through it. You are looking for a cohesive stretch with small tears, not translucency. See our whole wheat bread recipe for the full set of adjustments that whole grain demands.
High-rye dough: Rye contains almost no gluten-forming proteins. Structure in rye bread comes from pentosans (water-binding carbohydrates), not gluten. The windowpane test is irrelevant for rye percentages above 40%. Hamelman reduces mixing time dramatically for high-rye doughs — 100% rye gets 10-12 minutes on first speed only, with no second-speed mixing at all.
Very high-hydration dough (80%+): Ciabatta and other very wet doughs are too slack to stretch into a windowpane without tearing under their own weight. These doughs develop structure through extended folding during bulk fermentation rather than intensive mixing.
Enriched dough with high fat: Brioche dough at 50% butter will not windowpane the same way a lean dough does. The fat interferes with the stretch. Check for a smooth, glossy dough that pulls cleanly from the mixer bowl rather than a classic windowpane.
No-knead and minimal-intervention doughs: If you are following Forkish’s method (brief pincer mix plus folds), the dough develops through time and folding rather than intensive mixing. A full windowpane before bulk fermentation is not expected or necessary.
Common Windowpane Mistakes
Pulling too fast. The most common error. Yanking the dough tears it regardless of gluten development. Stretch slowly and gently, giving the gluten time to extend.
Using too small a piece. A piece that is too small is hard to stretch without tearing. Use a golf-ball-sized piece — large enough to grip comfortably with both hands.
Testing with very cold dough. Cold dough is stiffer and less extensible than warm dough. If testing after refrigeration, the window may fail even though the gluten is developed. Let the dough warm for a few minutes or test a piece that has been at room temperature.
Over-relying on the test. The windowpane tells you about gluten development at the moment of testing. It does not tell you about fermentation, hydration accuracy, or whether your leaven is active. It is one data point, not the whole picture.
The Windowpane in Context
The windowpane test fits into a broader system of dough evaluation:
| Stage | Test | What It Tells You |
|---|---|---|
| After mixing | Windowpane test | Is gluten sufficiently developed? |
| During bulk fermentation | Volume + jiggle | Is fermentation progressing at the right rate? |
| After shaping | Surface tension | Did shaping create a taut outer skin? |
| During proofing | Poke test | Is the dough properly proofed and ready to bake? |
| During baking | Internal temperature | Is the crumb fully set? |
Each test answers a specific question at a specific stage. Together, they give you control over the entire process. The windowpane is the first checkpoint — if gluten development is wrong, nothing downstream can fix it.
A Note on Overoxidation
Hamelman’s warning about overoxidation deserves special emphasis because it directly relates to the windowpane test.
The windowpane tells you when gluten is developed. It does not tell you to keep going until the window is perfect. If the dough passes the test — a thin, translucent sheet that holds together — stop mixing. Additional mixing beyond this point destroys carotenoid pigments (reducing flavor and color) and can eventually break the gluten bonds you just built.
This was the central insight of Raymond Calvel, the French baker who diagnosed the decline of French bread quality in the 1950s-60s. Industrial mixers were over-developing dough in pursuit of volume, destroying the very compounds that made bread taste like bread. Robertson trained under Calvel’s disciples and positions his entire method — minimal mixing, fold-based development — as a reversal of this mistake.
The windowpane test tells you “enough.” Respect that answer.
The Bottom Line
Tear off a piece, stretch it gently, and look for light. If you can see through it, your gluten is ready. If it tears, keep working. That is the entire test — five seconds, no equipment, clear signal.
It is the simplest and most useful technique check in all of bread baking. Use it every time.
Frequently Asked Questions
- What does the windowpane test show?
- The windowpane test shows whether your bread dough's gluten network is sufficiently developed. If you can stretch a small piece of dough thin enough to see light through it without tearing, the glutenin and gliadin proteins have formed enough cross-links to trap gas effectively during fermentation. If the dough tears before becoming translucent, it needs more mixing or folding time.
- How thin should the dough stretch for the windowpane test?
- Thin enough to see the shadow of your fingers through it, or even read text in strong light. Jeffrey Hamelman describes the target as 'a thin, almost transparent, sheet.' The key is translucency -- not paper-thin transparency. For fold-developed doughs (rather than mixer-developed), the window will be slightly thicker but should still hold together as a cohesive sheet.
- Can you over-mix bread dough past the windowpane?
- Yes. Once the dough passes the windowpane test, stop mixing. Additional mixing beyond this point can break the disulfide bonds holding the gluten network together, producing slack, sticky dough that cannot hold its shape. It also destroys carotenoid pigments, reducing flavor and color. Over-mixing is more common with stand mixers on high speed -- hand mixing rarely causes over-development because you would exhaust yourself first.
- Does the windowpane test work for whole wheat bread?
- Only partially. Bran particles in whole wheat flour physically puncture gluten strands, so whole wheat dough will never produce the thin, transparent window that white flour dough does. Instead, look for a cohesive stretch with small, clean tears rather than ragged shredding. If the dough pulls apart immediately with no stretch at all, it needs more development.
- Should I do the windowpane test for sourdough bread?
- It depends on your method. If you are using a stand mixer to develop gluten before bulk fermentation, yes -- check the windowpane after mixing. If you are using a fold-only method (stretch-and-fold during bulk), the windowpane is less relevant because gluten develops gradually over hours of folding. After 3-4 sets of folds, most doughs should produce a partial windowpane, which is sufficient.
