The forgetting curve is the sharp, predictable decline in memory that happens after you learn something new, first plotted by German psychologist Hermann Ebbinghaus in 1885. Left alone, most people lose the bulk of new information within the first day or two, then the rate of loss slows dramatically. The curve is not a fixed law of nature; how fast someone forgets depends on how the material was learned, how meaningful it is, and whether it gets reinforced afterward.
For L&D teams, the forgetting curve explains why a single training event rarely changes on-the-job behavior for long. This post covers what Ebbinghaus actually measured, what more than a century of replication and criticism has added to his data, and the specific tactics that measurably flatten the curve: spaced repetition schedules, retrieval practice, and reinforcement microlearning.
What Is the Forgetting Curve?
The forgetting curve is a graph showing how retention of newly learned information drops over time when there is no review. It falls fastest in the first hours and days after learning, then levels off into a long, shallow tail. Hermann Ebbinghaus plotted the first version of it in 1885 using himself as the only test subject.
Ebbinghaus expressed retention as a “savings score”: the percentage of original study time saved when relearning the same material later. A savings score of 34% at the one-day mark means relearning took 34% less time than the first attempt, not that a person could still recite 34% of the list from memory. That distinction matters, because L&D content often collapses the two into one loose idea of “retention.”
Who Was Hermann Ebbinghaus, and What Did He Actually Test?
Hermann Ebbinghaus was a German psychologist who, from 1880 to 1885, ran memory experiments on himself using lists of invented three-letter “nonsense syllables” such as WID and ZOF. He chose meaningless syllables on purpose, so no existing word association could help him remember them, isolating raw memory decay from prior knowledge.
He tested only himself, across thousands of lists, at intervals ranging from 20 minutes to 31 days, memorizing each list until he could recite it perfectly, then timing how long relearning took after a gap. He repeated the entire procedure a second time in 1883 and 1884 to check his own results. The single-subject design is a real limitation, but it is also why the data is so clean: no test material varied, and no other person’s motivation or background knowledge muddied the pattern.
How Fast Do People Really Forget New Information?
Ebbinghaus’s original data showed retention falling to about 44% after one hour, 34% after one day, and 21% after 31 days, with the steepest single drop happening inside the first hour rather than the first day. The decline then flattens into a long, slow tail rather than continuing at the same rate.
| Time Since Learning | Savings Score (Ebbinghaus, 1885) |
|---|---|
| 20 minutes | 58% |
| 1 hour | 44% |
| 9 hours | 36% |
| 1 day | 34% |
| 2 days | 28% |
| 6 days | 25% |
| 31 days | 21% |
These numbers come from nonsense syllables with no real-world meaning, so treat them as a shape rather than a promise that applies to your course content. A 2015 replication of Ebbinghaus’s forgetting curve reproduced this same pattern with a new subject and modern statistical methods, which is a rare confirmation for a 140-year-old psychology finding. Meaningful material, like a well-structured onboarding module tied to a learner’s actual job, tends to decay more slowly because it connects to things the learner already knows.
Is the “You Forget 70% of Training in 24 Hours” Stat Actually True?
No verified figure like “70% forgotten in 24 hours” appears anywhere in Ebbinghaus’s original data; it is a rounded, often-repeated paraphrase that training content has recycled for years without checking the source. Ebbinghaus’s real numbers show fast decay in the first day, but nowhere close to a flat 70% loss.
This is worth being precise about, because it happens to other well-known learning statistics too. A review of common forgetting-curve myths points out that people keep bolting invented percentages onto Ebbinghaus’s work the same way exaggerated numbers got attached to Edgar Dale’s Cone of Experience. What is real: a fast decline early on, followed by a slower one. What is not real: a single universal number that applies to every learner, every topic, and every training format. Treat any exact statistic quoted without a citation, including ones in articles like this, with some skepticism.
What Does Modern Research Say About the Ebbinghaus Forgetting Curve Today?
A 2015 replication by Jaap Murre and Joeri Dros reproduced Ebbinghaus’s curve almost exactly using a new subject and modern statistics, confirming the general shape while also showing that a simple exponential formula does not fit the data as well as more flexible models do. Their work also found that faster, deeper original learning predicts slower forgetting later.
The replication turned up something Ebbinghaus could not have measured with 1880s tools: retention around the 24-hour mark came in slightly higher than the smooth curve predicted, which the researchers linked to sleep’s role in memory consolidation. That detail matters for scheduling. A first review that lands after a night’s sleep tends to hold better than one crammed in a few hours later the same day. Researchers have also found that highly emotional or highly meaningful memories can deviate from the standard curve entirely, decaying much more slowly than neutral facts.
Why Does the Brain Let Go of New Information So Quickly?
New information fades fast because the brain treats it as provisional until repeated use signals that it is worth keeping. Without retrieval or reinforcement, the connections that encode a memory weaken through a mix of natural decay and interference from newer, similar learning that arrives afterward.
This is the same rewiring process covered in our guide to neuroplasticity and how your brain changes when you learn: connections that get used are strengthened, and ones that sit idle are gradually pruned. A brand-new memory starts out fragile and hippocampus-dependent, and only becomes stable and resistant to interference once it has been reactivated several times, ideally with sleep in between.
How Does Spaced Repetition Flatten the Forgetting Curve?
Spaced repetition works by inserting short review sessions right before a learner is likely to forget, which resets the decay clock and makes each later review last longer than the one before it. Over several cycles, the curve stops dropping toward zero and instead settles at a much higher plateau.
For the full mechanics of building a review schedule, including tools and spacing algorithms, see our guide to the spaced repetition study technique. The short version for L&D purposes: one exposure to new content is never enough, and the timing of the second exposure matters more than the timing of the third or fourth.
Match Spacing to How Often the Skill Gets Used
Give roles that use a skill daily a lighter reinforcement schedule than roles that touch it quarterly, since real-world practice is doing part of the spacing work for you already; applying one fixed refresher cadence across every role wastes reinforcement budget on people who don’t need it.
What’s the Best Spaced Repetition Schedule for Corporate Training?
There is no single universal schedule, but research on spacing gaps suggests reviewing material at roughly 20% of how long you need to remember it for retention windows measured in weeks, narrowing toward about 5% of the interval for content you need to retain for a year or more.
| How Long You Need It Retained | Research-Backed Gap Before First Review |
|---|---|
| 1 week | About 1 day after initial learning |
| 1 month | About 3 to 6 days after initial learning |
| 3 months | About 9 to 14 days after initial learning |
| 1 year | About 2 to 3 weeks after initial learning |
These gaps come from research on optimal spacing gaps, which mapped review timing against retention intervals ranging from days to years. Each subsequent review can then be spaced further apart than the one before it, since a stronger memory trace takes longer to decay.
How Does Retrieval Practice Slow Down Forgetting?
Retrieval practice, also called the testing effect, forces a learner to pull information out of memory instead of re-reading or re-watching it, and that act of retrieval strengthens the memory trace more than passive review does. This is why a short quiz beats replaying a training video, even when the video covers the same content more thoroughly.
This is the same mechanism behind active recall study techniques: the effort of generating an answer, even a wrong one followed by correction, does more for long-term memory than simply re-exposing a learner to the material. Retrieval practice and spaced repetition work best combined, since spacing decides when to test and retrieval practice decides what happens during that review.
What Is Reinforcement Microlearning, and When Should You Use It?
Reinforcement microlearning is a short follow-up activity, typically a two- to five-minute quiz, scenario, or flashcard set, delivered days or weeks after a formal course specifically to interrupt the forgetting curve before it settles at a low plateau. It is meant to reinforce, not re-teach.
Format matters here. A 2026 compilation of L&D statistics on training completion reports that short, three- to five-minute microlearning modules see completion rates two to three times higher than 30-minute courses covering the same ground. For reinforcement specifically, that length advantage matters more than content depth, since the goal of a reinforcement touchpoint is retrieval, not new instruction.
How Do You Calculate Training Retention Rate?
Retention rate is calculated by comparing a learner’s assessment score right after training to their score on an equivalent assessment given weeks later, expressed as a percentage: follow-up score divided by initial score, multiplied by 100. A learner who scores 90% right after training and 63% four weeks later has a retention rate of 70%.
Quiz-based retention rate is useful but incomplete, because multiple-choice recognition is easier than the unaided recall or applied performance the job actually requires. Where possible, pair a knowledge-based retention rate with a performance metric, such as error rate on the real task or time to complete it without a job aid, to see whether the training actually transferred.
Test the Task, Not Just the Quiz
At the follow-up checkpoint, ask learners to complete the real task without notes instead of retaking the original quiz; recognition-based retention numbers routinely run higher than actual on-the-job performance, so a quiz-only retention rate can mask a skill that has already faded.
Which Daily Habits Help Employees Retain Training Longer?
Small, consistent habits, such as reviewing one flashcard set at the same point in a daily routine, protect memory better than occasional long study sessions because consistency turns review into something automatic rather than something a learner has to remember to do.
Our guide to habit stacking for learning habits covers how to attach a short review to an existing routine, like checking email or starting a shift, so the review happens without relying on willpower or a calendar reminder. For L&D teams, this means building reinforcement into moments employees already return to daily, rather than expecting them to open a separate learning app on their own initiative.
How Should Instructional Designers Build Spacing Into Course Design From Day One?
Spacing should be part of the original course blueprint, not a reinforcement campaign bolted on after launch, because content designed for a single sitting rarely breaks apart cleanly into review-sized pieces later. Plan the first, second, and third review points before you finish the first draft of the course.
Encoding quality also affects how fast a piece of content falls down the curve in the first place. Content that requires a learner to explain a concept in their own words, the core idea behind the Feynman technique, tends to survive longer than content a learner only reads or watches passively, because explaining forces the same kind of retrieval effort that drives the testing effect.
Conclusion
Ebbinghaus’s forgetting curve is not an argument for giving up on training; it is a scheduling problem with a known shape. The curve tells you when memory is weakest, and spaced repetition, retrieval practice, and short reinforcement touchpoints tell you what to do about it.
Start with one course that matters, most compliance or safety programs are good candidates, and map three review checkpoints onto it using the spacing gaps above before you build anything new. Measure retention rate at each checkpoint with a real task, not just a quiz, so you know whether the schedule is actually working before you roll it out further.
If reinforcement is currently missing from your LMS workflow, closing that gap is usually the single most effective fix available, well ahead of redesigning the original course content.
FAQ
Q1. What is the forgetting curve in simple terms?
The forgetting curve is a graph showing how quickly people lose newly learned information without review. Memory drops fastest in the first hours and days after learning, then the rate of loss slows down and levels off. Hermann Ebbinghaus first plotted it in 1885 using memory experiments on himself.
Q2. Did Ebbinghaus really say people forget 70% of training in 24 hours?
No. That specific figure does not appear in his original data. Ebbinghaus’s actual numbers show retention falling to about 34% (measured as time saved relearning) after one day, using meaningless syllables, not real training content, which typically decays more slowly.
Q3. How is the forgetting curve different from the spacing effect?
The forgetting curve describes the problem: memory decays quickly without reinforcement. The spacing effect describes the fix: spreading review sessions out over time produces stronger, longer-lasting memory than cramming the same total amount of review into one sitting right before a test.
Q4. Does the forgetting curve apply to physical skills as well as facts?
The core pattern applies to skills too, though physical and procedural skills often decay more slowly than isolated facts because practice usually involves repetition and muscle memory. Skills still benefit from scheduled practice; long gaps without use lead to measurable performance drops.
Q5. How often should you use spaced repetition to counter the forgetting curve?
There is no single fixed interval. Research on optimal spacing suggests a first review at roughly 20% of how long you need to remember something, narrowing to about 5% of that interval for year-long retention goals, with each later review spaced further apart than the one before it.
Q6. Can the forgetting curve ever be fully stopped?
Not completely; some decay is a normal part of memory. But repeated retrieval and spaced review can flatten the curve so it settles at a much higher retention plateau instead of dropping toward zero, which is the realistic goal for most training programs.
