The Float Test: Does It Actually Work?

What the Float Test Actually Measures

The float test is a density check. Pinch off a small piece of starter, roughly 5 to 10 g, and drop it gently into a glass of room temperature water at around 68 to 72 °F (20 to 22 °C). If it floats, the theory says the starter is aerated with carbon dioxide and ready to leaven bread. If it sinks, it is judged flat and underfermented.

The mechanism is real. Active fermentation fills a starter with CO2 from bacterial and yeast activity. Those gas pockets lower the average density of the sample below the roughly 1.0 g/ml of water, so the piece rises. A flat, inactive sample holds little gas, stays denser than water, and drops to the bottom. The problem is that density depends on more than gas alone, and that is where the test starts to mislead.

How to Run the Test Correctly

Most float tests fail on technique, not on the starter. Use these conditions so the result reflects fermentation and not handling:

1. Small sample: a teaspoon, about 5 g, is enough. A large blob traps water in folds and behaves unpredictably.
2. Room temperature water: cold water increases the density difference slightly and is fine, but keep it consistent between tests so you are comparing like with like.
3. Gentle handling: lift from the surface, where gas concentration is highest, and set the piece on the water rather than dropping it from height.
4. Read it immediately: a floating piece will slowly absorb water and may sink after several minutes. The first 10 to 20 seconds is the honest reading.

Run under these conditions, a standard 100% hydration white flour starter at its peak will float reliably. That reliability is the reason the test spread so widely. It just does not survive contact with every kind of starter.

When the Float Test Is Unreliable

Density is set by trapped gas but also by the flour and the water content of the starter itself. Change either and the float result stops tracking readiness.

Whole grain and rye starters: whole wheat and rye are heavier and hold more bran, so the base mixture is denser before any gas is added. A rye starter fed at 100% hydration can be fully active at peak and still sink, because the flour weight overwhelms the buoyancy from CO2. Bakers reading that sink as failure feed again and lose a perfectly ripe starter.

Stiff and low hydration starters: a starter maintained at 50 to 65% hydration is thick, dense, and slow to trap large gas pockets. Many stiff levains stay active for hours yet never float, so the test returns a false negative on a starter that would bake beautifully.

High hydration starters: a slack starter at 125 to 166% hydration behaves more like batter. It can smear and cling rather than form a cohesive floating piece, and gas escapes as you pinch it off, producing a false sink.

Degassed samples: even an ideal 100% white starter fails the test if you stir it or scrape it hard while taking the sample. The act of removing the piece pushes out the very gas the test looks for.

The Over-Ripe Blind Spot

The float test has a second weakness that applies even to the white starters it suits. A starter that has peaked and just begun to collapse still holds a large volume of gas, so it continues to float for a while after it has passed its best point. The test reads "floating" as "ready," but it cannot distinguish a starter on the way up from one on the way down.

This matters because peak activity is a window, not a switch. A 100% hydration starter kept at 75 °F (24 °C) typically peaks 4 to 8 hours after feeding, then falls. Baking with a starter caught 2 to 3 hours after peak means weaker leavening and a flatter, more acidic loaf, even though the float test still shows green. Timing and appearance catch this. A cup of water does not.

Better Indicators of Starter Readiness

Every reliable readiness cue reads the starter directly rather than sampling a piece of it. Use these together:

• Rise: mark the level after feeding with a rubber band or tape. A ready starter has roughly doubled, and often risen 2 to 3 times, from that mark. This single cue works across all flours and hydrations.
• Bubbles: visible bubbles throughout the body and breaking on the surface, plus a webbed, aerated structure when you look at the side of a clear jar.
• Domed then flattening top: a rounded, domed surface signals peak. When the center starts to sink and flatten, the starter is just past peak.
• Aroma: a ripe starter smells yeasty and tangy, like beer or ripe fruit. A sharp acetone or nail polish remover smell means it is hungry and overdue for a feed.
• Timing: track how many hours your starter needs to peak in your kitchen. Once you know it peaks in, say, 6 hours at 74 °F (23 °C), the clock becomes your most repeatable tool.

Where the Float Test Really Shines: Dough, Not Starter

The float test is more useful for judging bulk fermentation of a shaped-in dough sample than for judging starter reripeness, though its clearest use is a variant applied to dough. When bulk fermentation is complete, a well-aerated dough is buoyant. Pinch off a walnut-sized piece near the end of bulk and float it: a piece that holds together and floats signals the dough is gassy and ready to shape, while a dense piece that sinks suggests bulk needs more time. Because the whole dough is at a consistent hydration you already know, the density reading is less confounded than it is for a spoonful of varied starter. Pair this with a tool like the bulk ferment timer to estimate how long your dough needs at its current temperature, then confirm with the float and the poke test rather than relying on any single check.

The Verdict

The float test is a nice-to-have quick check, not a decision maker. For a 100% hydration white flour starter handled gently and read immediately, a float is a fair sign of activity. For whole grain, rye, stiff, or very slack starters, it produces false sinks that send bakers into needless refeeding. And even at its best it cannot tell a rising starter from a collapsing one.

Lean on rise, bubbles, a domed top, aroma, and consistent peak timing to judge starter readiness, and save the float test for confirming that a shaped dough has built enough gas at the end of bulk. Treat it as one input among several, and it stops lying to you. The final proof of a starter is always the loaf: an active, bubbly starter that sank in the glass can still deliver an open, well-risen bread.