Six visual signs, the float test reality check, and why the clock alone gets it wrong. Plus a free app that predicts peak time based on your temperature and feeding ratio.
Or try the web version →If you've ever fed your starter, gone to work, and come home wondering "is it ready or did I miss the peak?" — this guide answers exactly that. There are six visual signs that tell you. The float test sort of works. And the clock alone reliably lies. Here's how to read your starter, and how a quantitative schedule turns "wait and check" into a planned process.
A sourdough starter is ready to use when it has tripled in volume from its post-feed level, the surface is domed and just starting to flatten (not yet receding), the jar sides show active webbed bubbles, and the aroma is yeasty with a clean mild tang (not sharp vinegar). This is "peak" — typically a 30–60 minute window. Past peak, the starter still works but produces less rise and more acidic flavor.
The single most reliable sign. Mark your starter's level with a rubber band immediately after feeding. Peak is when the culture reaches 3× that mark with a smooth dome that has just stopped rising. If you see a 4× dome still climbing, you're 30–60 minutes pre-peak. If you see a slight depression forming, you're a few minutes past peak.
A pre-peak starter has a dome that's still climbing. At peak, the dome plateaus — perfectly smooth, no movement. Past peak, the center starts sinking into a crater. Severe past-peak shows the top "cliff-collapsed" against the jar walls.
Look at the side of the jar. At peak, you see a dense network of small-to-medium bubbles, all "alive" and contained. Past peak, the bubble walls collapse into larger irregular voids with shiny streaks where gas has escaped.
A ready starter smells primarily of bread — yeasty, slightly sweet, with a clean lactic undernote. No acetone. No nail polish remover. No sharp vinegar punch. A vinegar-dominant smell means significantly past peak.
Press your finger lightly into the surface. At peak, the indentation springs back about halfway and the texture feels light but slightly resistant. Pre-peak: dense, resists the press. Past peak: indentation stays, texture has gone slack.
Bring the jar near your ear. At peak you can often hear soft irregular pops as bubbles break the surface. Most subjective of the signs, but the most fun.
The float test — drop a spoonful into water and see if it floats — is the most widely cited test in home sourdough, and the most overstated. It works on a simple principle: if the starter is full of trapped gas, it floats.
Where the float test passes: Anywhere from about 70% rise through peak and 1–2 hours past peak. That's a wide window. A passing float test tells you "the starter has gas in it." It doesn't tell you "peak is now."
Where the float test fails at peak: Whole-grain or rye starters at high hydration sometimes don't float at peak because their gas-bubble distribution is different. Low-hydration stiff starters (50% water) often don't float regardless of activity. Some healthy starters with very fine, dense bubble structures don't float well even at peak.
The practical rule: Use the float test as a binary sanity check ("did this thing rise and trap gas? yes/no"). Use volume, surface, and aroma to locate peak itself.
"Feed your starter 6–8 hours before mixing" appears in countless recipes. It's right sometimes. It's wrong often. Here's why:
| What changes | Peak-time impact |
|---|---|
| Room temperature 22°C → 26°C | Peak time cuts ~30% |
| Room temperature 22°C → 18°C | Peak time extends ~50% |
| Feeding ratio 1:5:5 → 1:1:1 | Peak time cuts ~60% |
| Feeding ratio 1:5:5 → 1:10:10 | Peak time extends ~80% |
| White flour → whole wheat | Peak time cuts ~33% |
| White flour → rye | Peak time cuts ~50% |
A "6 hours" rule of thumb might land you at 70% rise (under-ready) or 2 hours past peak (over-acidic) depending on which way the variables drift. The fix isn't ignoring the clock — it's adjusting the clock for your actual conditions.
Yeast and bacteria activity scales exponentially with temperature (Q10 ≈ 2 — rate doubles per 10°C). It scales non-linearly with inoculation ratio (lower ratio = longer lag + slightly longer exponential phase). It scales per-flour-type (whole grains have more amylase and minerals → faster fermentation).
Here's a reference table for a "standard" white-flour starter at common conditions:
| Feed Ratio | 18°C peak | 22°C peak | 26°C peak | 30°C peak |
|---|---|---|---|---|
| 1:1:1 | 4–6 h | 2–4 h | 1.5–3 h | 1–2 h |
| 1:2:2 | 5–8 h | 3–5 h | 2.5–3.5 h | 1.5–2.5 h |
| 1:5:5 | 10–14 h | 6–8 h | 4–5 h | 3–4 h |
| 1:10:10 | 14+ h | 10–14 h | 7–10 h | 5–7 h |
These are estimates. Your specific starter, flour, and storage conditions will produce some variation. Use these as starting points — refine over a few bakes.
Input your target time, temperature, ratio, and flour. The app tells you exactly when peak will occur and when to feed. Free for iOS, Android, and web.
Two good options. Wait — another 30–90 minutes at room temperature usually gets you to peak. Or refrigerate — cold pauses the process. You can hold a near-peak starter at 4°C for 12–18 hours; when you bring it back to room temperature, peak resumes within 1–2 hours.
You can still bake with it. Expect slower rise (around 30% less than peak-fresh) and more sour flavor. Best for sourdoughs where you want pronounced tang anyway.
Two options. Feed it again at the same ratio and wait 2–4 hours for the next peak. Or use it as-is but add 10–20% commercial yeast as a safety net for rise.
This starter is too acidic for a same-day bake. Do two feeds over 24 hours (e.g. 1:5:5 in the morning, 1:5:5 in the evening, use the third feed the next morning) to restore yeast population and lower acidity.
Every starter has a personal rhythm shaped by its strain mix, your flour, your storage habits, and your kitchen temperature. Track yours over 4–7 consecutive feeds at a consistent ratio and temperature. The average peak time tells you your starter's natural cadence at that ratio + that temperature.
Once calibrated, extrapolate: doubling the ratio (1:5:5 → 1:10:10) roughly doubles time. Halving (1:5:5 → 1:2:2) cuts time by 30–40%, not 50% — because lag and exponential phase shift non-linearly.
Combine your calibrated cadence with a target-time calculator and you stop guessing. You set when you want bread, and the math tells you when to feed.
Tripled in volume from post-feed, smooth domed surface just starting to flatten, webbed bubble structure on jar sides, yeasty + mild tang aroma, slight springiness when pressed. All five together = peak. Use within the next hour for best results.
Triple is the target for a 1:5:5 feed at 22°C. For 1:1:1 feeds, expect 4–5× rise. For 1:10:10, expect 2.5–3×. The exact multiplier depends on inoculation — but for any specific feed ratio, hitting peak is more important than the exact multiplier.
1:1:1 at 22°C: 2–4 hours. 1:2:2: 3–5 hours. 1:5:5: 6–8 hours. 1:10:10: 10–14 hours. Each interval shortens about 30% per +4°C and extends about 50% per −4°C.
Yeasty, slightly sweet, mildly tangy. Like bread dough with a clean lactic note. NOT sharp vinegar, NOT nail polish remover. If it smells like vinegar, it's past peak. If it has alcohol notes (hooch on top), it's significantly past peak.
Don't panic. The float test fails on high-hydration whole-grain starters and on stiff (low-hydration) starters even at peak. Trust the volume + surface + aroma signs instead. A dome-shaped, tripled, mild-tangy starter is ready whether or not it floats.
Yes. Refrigerated at 4°C, a peaked starter holds within a usable window for 12–18 hours. After that the acidity continues building slowly. Bring back to room temperature; use within 1–2 hours of warming.
Surface depression (crater forming), sharp acetic aroma replacing the mild tang, slick or glossy sides where bubbles have collapsed, and a slack feel when pressed. All four = significantly past peak. Feed again or use with caveats.