The Forgetting Curve: What Ebbinghaus Found, and How to Beat It

The forgetting curve shows how quickly we lose something new when we do not go back to it. Memory falls fast in the first hours and then much more slowly. Hermann Ebbinghaus measured it on himself in 1879 and 1880, and his result has been repeated since. This guide shows what he actually found, what the curve does and does not mean, and how to flatten it.

Quick Answer

1. The curve is steep, then flat. Most forgetting happens in the first hours after learning.
2. It is not all lost. A month later, Ebbinghaus could still relearn a list about a fifth faster than the first time.
3. Reviews flatten it. Each time you recall something successfully, the next fall is slower.
4. So review at growing gaps: after a day, a few days, a week, then a month.

What Is the Forgetting Curve?

The forgetting curve is a graph of memory over time. On the left, just after you learn something, you know it well. As time passes, the line drops: first steeply, then more and more gently. It never quite reaches zero, because some trace of the learning stays.

The curve describes memory with no review at all. That is the important part. The shape changes as soon as you go back to the material, which is why the curve is the starting point for every good study method.

How Ebbinghaus Measured Forgetting

Hermann Ebbinghaus was a German psychologist who wanted to measure memory as carefully as a physicist measures weight. He was his own and only subject. To keep meaning from helping him, he learned lists of nonsense syllables such as WID and ZOF, with 13 syllables in a list.

He read each list aloud at a steady pace until he could recite it twice without a mistake, and he timed how long that took. Then he waited: 20 minutes, an hour, 9 hours, a day, two days, six days or 31 days. After the wait, he learned the same list again and timed it once more.

The savings method

If the relearning was faster, some memory must have been left. Ebbinghaus called the time saved the savings.

This is a clever measure. Even when you cannot recall a single syllable, relearning can still be faster, so savings pick up memories too faint to bring back on their own.

What He Found

Here are Ebbinghaus's results beside a modern repeat of the experiment from 2015. Both follow the same shape: a steep fall in the first hour, then a long, slow decline.

Savings in relearning time. Ebbinghaus's figures are from his book; the replication is Murre and Dros (2015), both as printed in Table 3 of their paper.

Put into plain words, for Ebbinghaus:

• After one hour, relearning took 56% of the first effort.
• After one day, it took 66%. In his own words, "about one third was always remembered."
• After a month, it still took only 79%: "fully one fifth of the first work persisted in effect."

What the Forgetting Curve Does Not Say

The curve is often shared as "you forget 50% in an hour and 70% in a day". Those numbers do not come from Ebbinghaus's results, and they mix up two different things.

• Savings is not recall. Ebbinghaus measured how much faster relearning was, not how many items he could remember. A savings of 34% after a day means relearning took two thirds of the time, not that 66% of the list had vanished.
• Nonsense is the worst case. He chose meaningless syllables on purpose, so that nothing he already knew could help. Material you understand, and can link to things you know, is usually easier to keep.
• It is one person's curve. The exact numbers depend on the person, the material and how well it was learned. The shape is what holds, not the precise percentages.

So use the curve as a picture of how forgetting behaves, not as a promise of how much you will lose by tomorrow.

Has It Been Replicated?

Yes, more than once. The best-known repeat was published in 2015 by Jaap Murre and Joeri Dros. Dros was the only subject, as Ebbinghaus had been, and learned 70 lists of nonsense syllables over about 70 hours, with the same seven intervals. His curve closely matched the original. The authors concluded that the forgetting curve "has indeed been replicated and that it is not completely smooth."

The "not completely smooth" part is about the points at one and two days. In every data set the authors looked at, at least one of them sat a little above a smooth curve. In the chart, Dros's savings even go up, from 27.6% at nine hours to 31.7% at one day. The authors suggest sleep as a possible reason, but their statistical test found only a trend, not a clear effect. The biggest difference from the original is at 31 days, where Dros's savings were much lower than Ebbinghaus's.

How to Beat the Forgetting Curve

You cannot stop forgetting, but you can change the curve. Four things have strong support.

1. Review at growing gaps

Each review lifts your memory back up, and after a successful one the next fall is slower. So the gaps between reviews can grow. This is called spaced repetition, and it is one of the most tested findings in psychology. A 2006 review pulled together 317 experiments from 184 articles on spacing.

How long should the gaps be? It depends on how long you need to remember. In a 2008 study of more than 1,350 people, the best gap before a review was about 20–40% of the time until the test when the test was a week away, and about 5–10% when it was a year away. Simply put: the longer you need to remember, the longer the gaps can be.

A starting schedule, not a rule. If you miss something, bring that item back sooner. If it is easy, let the gap grow.

2. Test yourself instead of rereading

A review works best when you pull the answer out of your memory before you look. In a 2006 study, students who took recall tests on a passage remembered much more two days and a week later than students who reread it the same number of times, even though the rereaders felt more confident. See active recall for how to do it.

3. Learn it in a form that lasts

Ebbinghaus used nonsense to make memory as hard as possible. You can do the opposite. Turn what you learn into vivid pictures and place them along a route you know, as in the memory palace. In a 2017 study, people with ordinary memories who trained in this method for six weeks recalled far more words, and kept much of the gain four months later.

4. Sleep on it

The bump at one day in the replication is only a hint, and for this kind of learning the effect of sleep is not settled. Still, a short review in the evening and another the next morning costs almost nothing, and it puts your first review right on the steep part of the curve.

The Forgetting Curve and Memory Athletes

In a memory championship, recall comes within minutes or an hour of learning, so athletes work at the steep top of the curve. Their images have to be strong enough to survive that first drop. Outside competition, athletes keep their systems, such as their Major System words, with spaced reviews so the pictures come back instantly.

MemoChamp's Review does this for the words and images you build: the ones you find hard come back sooner, and the ones you know drift further out. For a long number, the Pi Trainer lets you check how much of pi survived since your last review.

Frequently Asked Questions

Core Takeaways

1. Without review, memory falls fast in the first hours and then slowly.
2. Ebbinghaus measured savings in relearning time, not the share of items recalled.
3. The popular "70% gone in a day" figure is not what his data show.
4. A 2015 replication found the same shape, with a possible boost from sleep.
5. Spaced review, self-testing and vivid images flatten the curve.

Related Guides

• Spaced Repetition — how to build a review schedule that adapts to you.
• Active Recall — why testing yourself beats rereading.
• Method of Loci — learn material in a form that lasts.
• Memory Training for Beginners — a four-week plan with reviews built in.

Sources

Read on 3 November 2026: Hermann Ebbinghaus, Memory: A Contribution to Experimental Psychology, chapter 7 (1885; translated by Ruger and Bussenius, 1913); Murre and Dros, "Replication and Analysis of Ebbinghaus' Forgetting Curve," PLoS ONE (2015); Cepeda et al., "Distributed practice in verbal recall tasks," Psychological Bulletin (2006); Cepeda et al., "Spacing effects in learning: a temporal ridgeline of optimal retention," Psychological Science (2008); Roediger and Karpicke, "Test-enhanced learning," Psychological Science (2006); Dresler et al., "Mnemonic training reshapes brain networks to support superior memory," Neuron (2017).