BakeHQ

The loaf burst or split at the side

The loaf opened along an unplanned line during baking — commonly where a braid or seam ran, or around the base where the crust met the tin — leaving a ragged tear paler than the crust around it.

Usually established at: fermenting

Narrow it down

Answer what you can. Anything you skip simply leaves its causes in play — there is no requirement to answer everything, and no answer is treated as more authoritative than what you actually saw.

What did the shaped loaf do when pressed just before it went in?
Was it covered while it proved?
What does the inside of the tear look like?
Was the loaf scored before baking?

All 4 possible causes are shown below. Answer a question above to start ruling them out.

Every possible cause

The complete model, whatever you answered above. Each cause states the mechanism — what physically happens — because a cause without a mechanism is folklore with a confident tone.

The loaf went in with expansion still to come

Very common

A loaf loaded before fermentation is finished still has gas to produce and dough that can still expand. That expansion happens in the oven, against a surface that is drying and setting from the outside in. Where the crust has set first, the expanding interior has to tear it to get out — so the loaf bursts along whatever line was weakest, usually a seam, the side, or the base of a braid. THE SIGNATURE IS THAT THE TEAR IS RAGGED AND PALE INSIDE, because it opened after the surface had already coloured.

What to change

  • Prove until the dough says so, not until the clock does

    Judge the second prove by the dough: a floured finger pressed lightly into the shaped loaf should leave a dent that fills back slowly and incompletely. If it springs straight back it is not ready, whatever the recipe's time says. A cool kitchen can easily double the time a recipe assumes.

    Shaped and proved yeasted loaves. It does NOT apply to a loaf that was already at or past peak, where more time makes the collapse worse rather than better.

    Trade-off: The window between ready and overproved is real, so this is a check to repeat rather than an instruction to leave it longer.

Underproofed

Very common

Fermentation rate depends strongly on temperature, and a dough in a cool kitchen takes considerably longer than the same dough in a warm one. A recipe’s stated time describes the kitchen it was written in. Baked early, the dough has not accumulated enough gas, and the structure sets around a smaller volume.

The variable underneath

  • Dough temperature

    How to tell after the fact: A loaf that over-proved in less time than the recipe stated, in a warm kitchen, indicates dough that started warm.

What to change

  • Judge proofing by the dough, not the clock

    Press a floured fingertip about a centimetre into the dough. If the indentation springs back fully and quickly it needs longer. If it springs back slowly and partly, it is ready. If it does not spring back at all and the surface may sag, it has gone too far.

    Any yeasted or naturally leavened dough at final proof. Less informative during bulk fermentation, where volume and surface condition are better guides.

    Trade-off: Requires handling the dough, which slightly disturbs the surface. This is immaterial compared with baking at the wrong point.

  • Control the dough temperature via the water

    Adjust the temperature of the mixing water to bring the finished dough to a consistent target. Warmer water in a cold kitchen, cooler water in a warm one. Measure the dough after mixing so the adjustment can be repeated.

    Any fermented dough. Particularly valuable for sourdough, where timing is otherwise unpredictable.

    Trade-off: Requires a thermometer to do properly, and a note of what worked last time.

The surface dried into a skin before it went in

Common

Dough left uncovered in moving air loses water from its surface and forms a skin that is measurably less extensible than the dough beneath it. In the oven the interior expands normally and the skin does not, so the loaf splits where the skin gives way. It is the same visible outcome as a loaf that was underproved, arrived at from the opposite direction — this dough may have been proved perfectly and simply dried while it did.

What to change

  • Keep it covered while it proves

    Cover the shaped loaf with a box, an upturned bowl or oiled wrap for the whole of the second prove — anything that stops air moving across the surface. A damp cloth laid directly on an enriched dough can stick to it; a cloth over a support does not.

    Every shaped prove, and especially near an oven, a radiator or a fan, where a surface dries fastest.

    Trade-off: None worth stating, beyond remembering to take the cover off before loading.

The loaf was not scored, or was scored too shallowly

Common

In the first minutes of baking the surface sets while the interior is still expanding. A score is a deliberate weak point that gives that expansion somewhere to go before the crust is rigid. Without one, the loaf bursts at whatever the weakest part of the surface happens to be — often a side seam, sometimes underneath — and the visible result is a split flank or a loaf that has simply stopped expanding because it had nowhere to open. A score that is too shallow does nothing, because the set crust is thicker than the cut. This applies to free-standing loaves and much less to tin loaves, which are supported on three sides and expand upwards regardless.

What to change

  • Score decisively, or accept where the loaf chooses to open

    Use a blade rather than a knife, cut in one confident movement, and make the cut deep enough to be a genuine weak point rather than a mark on the surface. One decisive cut does more than several tentative ones.

    Free-standing loaves. Tin loaves and enriched breads baked in moulds need it far less, and some breads are deliberately left unscored so that they tear.

    Trade-off: A deep score on an over-proofed loaf can help it deflate rather than help it open, so scoring does not compensate for proof state. It directs expansion that exists; it does not create it.

Preventing it next time

  • Prove the shaped loaf until a pressed finger leaves a dent that fills back slowly and incompletely — not until a stated time has passed.
  • Keep it covered for the whole of the shaped prove, so the surface never dries into a skin that cannot stretch.
  • Give the expansion somewhere to go: score decisively, or accept that a braided loaf will open along a seam and shape it so that seam is underneath.

A burst loaf is the opposite problem to a flat one

It is worth saying at the top, because the two faults are usually confused and their corrections are opposite.

A loaf that stays FLAT had nothing left to give: the gas was spent, the network was degraded, or it never fermented at all. More time in the bowl helps it.

A loaf that BURST had plenty left. It expanded hard enough to tear its own crust open, which is a demonstration of strength rather than weakness. More time in the oven does nothing for it, and more yeast makes it worse.

So the question is never “did it rise enough” — the loaf has already answered that. It is where the rise was allowed to go.

Underproved and skinned tear the same way, and one of them proved perfectly

Both produce a ragged pale tear, and the discriminating question is not about the bread at all.

An UNDERPROVED loaf still had fermentation to finish. It goes into the oven with the dough capable of a great deal of expansion, the yeast accelerates in the heat before it dies, and the surface sets before the interior has finished moving. The tell is the finger test before loading: the dent vanishes instantly.

A SKINNED loaf may have been proved exactly right. It dried on the outside while it proved — twenty minutes uncovered near a warm oven will do it — and a dry skin is simply less extensible than the dough under it. The interior expands normally, the skin will not follow, and it splits. The tell is that the dent behaved correctly and the loaf was uncovered.

The corrections do not overlap. One is more time; the other is a box over the top and no more time at all.

Scoring decides where it opens, not whether it opens

A scored loaf opens along the cut because that is the weakest line in the crust, and the baker chose it. An unscored one opens wherever the crust happens to be weakest, which is usually a seam, a braid junction or the join at the base.

So scoring does not prevent this fault. It relocates it, and relocating it is most of what a reader wants — an opened score reads as a loaf, a burst side reads as an accident.

What scoring cannot do is absorb the expansion of a loaf that went in badly underproved. That one will tear somewhere in addition to the score, and the extra tear is the evidence.

Braided and enriched loaves are the common case in this catalogue precisely because they are usually unscored: a challah is not cut before baking, so every bit of its expansion has to leave through a seam.

Related

Where this happens

  • Challah

    Braided and never scored, so every bit of expansion has to leave through a seam — which is exactly where it tears.

  • Brioche

    Enriched dough ferments slowly and is routinely loaded before it is ready, because the clock says it should be.

  • Bread rolls

    Small units skin fastest, and a tray of rolls is usually proved uncovered on the worktop.

  • Hot cross buns

    Heavily enriched and fruited, so the prove is long, and the piped cross gives the surface a second line to fail along.

  • Milk bread

    Baked in a tin and unscored, so it splits where the crust meets the rim.

Possible causes

  • The loaf went in with expansion still to come

    The finger test before loading separates this from every other cause here, and it takes two seconds.

  • Underproofed

    The general model behind it: a recipe’s stated time describes the kitchen it was written in, and a cool kitchen doubles it.

  • The surface dried into a skin before it went in

    The case where the prove was correct and the loaf tore anyway, which is why it is asked about separately.

  • The loaf was not scored, or was scored too shallowly

    Decides WHERE a loaf opens rather than whether it does — which is most of what the reader is complaining about.

Commonly confused with

  • Little or no oven spring

    The two opposite readings of the same loaf. A weak spring means the dough had nothing left to give and the answer is usually more fermentation; a burst side means it had a great deal left and more fermentation is exactly wrong. The finger test before loading separates them in two seconds.

  • The loaf collapsed or sank

    A loaf that collapsed went past its peak; one that burst never reached it. Both look like a loaf that failed in the oven, and the correction for either makes the other worse.

If none of this matches what you saw, the model has a gap. Tell us — that is how this section gets better.

Keep reading

  • Challah

    What the thing is supposed to be like, which is where most diagnoses actually start.

  • Why baking fails

    The small number of things that go wrong underneath the large number of ways it shows.

BakeHQ would like to record anonymous usage — which pages are read, whether searches return anything, whether a diagnosis reaches an answer. No personal data, no advertising profile, and nothing you type is ever sent. The site works identically either way.