← Blog
2026-09-05

Your starter is bubbling but not rising

Bubbles mean your starter is alive and producing gas. If the level in the jar never climbs, the gas is escaping instead of lifting, and the most common reason is that you're looking at the jar after it already peaked.

Your starter is fine. The bubbles prove it.

Bubbles are carbon dioxide. Carbon dioxide in a flour-and-water paste means yeast is eating and respiring in there. A dead starter does not bubble. Whatever is going wrong, it is not that you killed it, and it is not that you need to throw it out and start over.

The rise is a separate question from the life. Gas production and gas retention are two different things, and only one of them shows up as a line climbing the side of the jar.

What you're seeing

You feed the starter. A few hours later the surface is domed with bubbles, or the whole mass is shot through with them like a sponge. It smells yeasty, or sharp, or like nail polish remover. You look at the jar and the level is exactly where you left it, or close to it. Maybe a fingerwidth up, nothing like the doubling every video promises.

Some people see this at every feed for weeks. Some see it on day 4 of a brand-new starter and panic.

The cause: you're looking after the peak

This is it, the overwhelming majority of the time. Your starter rose. Then it fell. And you walked back into the kitchen somewhere on the far side of that curve and saw a flat, bubbly jar.

A starter's rise is a curve, not a plateau. The yeast consumes the available sugars, inflates the paste with CO₂, and reaches a peak. At the peak the gluten and the fermentation byproducts can no longer hold the structure up, the network breaks down, and the whole thing sinks back to roughly where it started. What's left behind is a flat, heavily bubbled jar. The exact thing you're staring at.

A vigorous starter in a warm kitchen at 78°F (26°C) can complete that entire cycle in 4 to 6 hours. If you feed at 9 p.m. and check at 7 a.m., you missed it by hours and there is no evidence remaining except bubbles.

Did I miss the peak?

fedpeak 6h 36m · 74°F0h6h12h18h24hhours after feeding, at 74°F (23°C)
74°F / 23°C
60°F / 16°C85°F / 29°C
10h after feeding
straight after feeding24h

You missed it. This is the fall.

Nothing is wrong with your starter. It rose to double at 6h 36m after feeding at 74°F (23°C), then the network gave out and it sank back. At 10h you are looking at the wreckage: flat, heavily bubbled, and still perfectly alive. Feed it and check at 6h 36m next time.

Can't sit and watch the jar? Snap a rubber band around it at the level of the starter right after feeding. The residue clinging to the glass above the band records the high-water mark, so a peak you slept through still gets measured.

The other tell: hooch. A thin layer of dark liquid on top, or a strong acetone smell, means the starter ran out of food some time ago. That is a post-peak starter, not a weak one.

The fix: catch the peak

Stop checking whenever you happen to walk by. Measure it.

  1. Rubber band the jar. After feeding, snap a rubber band around the jar at the level of the starter. Now the rise is recorded whether you're watching or not, and so is the fall, because the residue clinging to the glass above the band shows you the high-water mark.
  2. Use a narrow, straight-sided container. A wide bowl spreads 50 g (1.8 oz) of starter into a thin layer where a doubling is barely visible. A narrow jar makes the same rise obvious.
  3. Float a sample. Put a tablespoon of freshly fed starter in a small shot glass or narrow jar alongside the main one. It's a low-effort proxy that's easier to read.
  4. Check every hour for one cycle. Once. Write down when it peaks at your kitchen's temperature. Then you know your starter's schedule instead of guessing at someone else's.

The other three causes, briefly

Hydration too thin. A starter fed to a batter or pancake-mix consistency can't trap gas. The bubbles form, rise through the slurry, and pop at the surface. You get foam instead of lift. Feed at 100% hydration by weight, equal grams of flour and water: 50 g flour (1.8 oz, about ⅓ cup) to 50 g water (1.8 oz, 50 ml). It should be a thick, spoonable paste that mounds, not pours.

The flour. All-purpose bleached flour is the worst-performing starter food in a home kitchen: bleaching strips the flour of some of what the culture feeds on, and low-protein AP builds a weaker gas-holding network. Switch to unbleached bread flour, or add 20% whole wheat or rye to every feed. Rye is the fastest fuel there is.

Too cold. Below about 68°F (20°C) fermentation slows dramatically. A starter that doubles in 5 hours at 78°F (26°C) may take 12 or more at 65°F (18°C), which reads as "it never rose" if you check it on a normal schedule. Find a warmer spot: an oven with just the light on, the top of the refrigerator, a proofing box.

Starter temperatureRoughly when to expect peak
65°F / 18°C10–14 hours
70°F / 21°C8–10 hours
75°F / 24°C6–8 hours
80°F / 27°C4–6 hours

Those are ranges for an established starter at 100% hydration with a 1:1:1 feed. Your starter's number is the one you measure, not the one in the table.

If your starter is less than two weeks old

Different situation, same non-emergency. A brand-new starter usually bubbles enthusiastically on days 2 and 3, and that early activity is mostly bacteria, not the yeast you actually want. Around days 3 to 5 that population dies back as the environment acidifies, and everything goes quiet. Flat. Sometimes for several days.

That stall is normal and it self-corrects. Keep feeding on schedule. The real yeast establishes itself in the acidic environment and takes over, usually by day 7 to 14. Do not throw it out on day 5.

Is it saveable?

There is nothing to save. A bubbling starter is a working starter.

If you want to make it stronger before you bake with it, feed it twice a day for three days at a 1:1:1 ratio by weight (50 g starter, 50 g flour, 50 g water, 1.8 oz each), using unbleached bread flour with 20% whole rye, kept at 75–80°F (24–27°C). Three days of that turns a sluggish culture into a reliable one.

The test that actually matters: does it at least double from a fresh feed, on a predictable schedule you've measured? If yes, bake with it at its peak. If it never doubles after a week of consistent warm feeding, then it's the flour or the temperature, in that order.

Something else wrong?

Or start at the diagnosis index.

← All posts