Spacing effect is stranger than it sounds.

Studying for three hours straight is nearly the worst way to study. Ten minutes a day for eighteen days beats it.

The finding is called the spacing effect, and it is one of the most replicated results in the science of learning. Take any material, vocabulary, anatomy, physics, and compare two schedules with equal total study time: massed practice, cramming it all into one session, versus distributed practice, spreading the same minutes across days or weeks. The spaced schedule wins on every delayed test, often by a large margin. Cramming can win on a test given immediately, which is exactly why students keep doing it, and then the knowledge evaporates.

Hermann Ebbinghaus stumbled onto the effect in the 1880s during his heroic self-experiments on memory. Memorizing lists of nonsense syllables, he noticed that spreading his repetitions across days let him learn in fewer total repetitions than hammering them all at once. He was documenting the forgetting curve at the same time, and the two discoveries belong together: memory decays fast at first, and each spaced re-encounter interrupts the decay and flattens the curve a little more.

Why does spacing work? Several mechanisms stack up. Each spaced encounter requires effortful retrieval, you have to dig the half-faded memory back up, and that effort strengthens the trace. The encounters also happen in slightly different contexts, different moods, different rooms, giving the memory more retrieval routes. And spacing defeats the illusion of fluency: when you reread notes in one sitting, the material feels familiar, and familiarity masquerades as knowledge. Spaced out, the familiarity fades between sessions, so you get an honest signal of what you actually retain.

The modern evidence is overwhelming. A landmark 2006 review by Cepeda and colleagues, analyzing hundreds of studies, confirmed that spaced practice beats massed practice across subjects, ages, and time scales, and even mapped the optimal gaps: for long-term retention, the space between sessions should be a meaningful fraction of the time until the test. The research has a name for the underlying principle, desirable difficulty: learning that feels harder in the moment, retrieval with effort, spacing with forgetting in between, produces stronger memory than learning that feels easy.

The gaps themselves have an optimal size. Cepeda and colleagues followed their 2006 review with work mapping the ideal spacing interval, and the answer is elegantly proportional: the best gap between study sessions is roughly ten to thirty percent of the retention interval. Remembering something for a week? Space sessions about a day apart. For a year? Space them weeks apart. Too short a gap and the retrieval is effortless, building little. Too long and the memory is gone beyond retrieval. The sweet spot is the edge of forgetting.

A close cousin is interleaving. Where spacing distributes repetitions of one topic over time, interleaving mixes different topics within a session: instead of drilling twenty identical algebra problems, alternate problem types. Rohrer and Taylor showed in 2007 that interleaved practice dramatically improved later test performance in mathematics, because it forces the brain to practice choosing the right approach, not just executing it. Spacing and interleaving together are the closest thing learning science has to a free upgrade.

The practical version fits on an index card. Study in short sessions. Leave gaps. Let yourself partly forget, then retrieve. Flashcards beat rereading. Sleep between sessions. The method feels slower and less productive than the highlighter-and-cram, and that feeling is the oldest trap in the literature: the strategies that feel effective are often the ones that fail, and the one that feels like struggling is the one that works.