Does Cramming Work? The Short Answer
It works better than not studying, and that's where its advantage ends. A long session the night before will raise tomorrow's score compared with doing nothing. It won't beat the same hours spread over several days: in Nate Kornell's flashcard study, words crammed on the last study day were recalled at 34% the next day, against 54% for words whose study was spread across four days. In experiments that test people both soon and later, massed study comes out ahead only on a test given minutes after studying.
The bigger problem is how fast crammed material fades. When the same topics come back on the final, you end up learning them twice. And if the cram turns into an all-nighter, you also cut into the sleep your brain uses to keep what you studied.
So cramming is a workable way to scrape through tomorrow and a poor way to still know the material next month. If you're reading this the night before an exam, the step-by-step plan further down is a better way to cram than the one most people use.
What Cramming Means in Research Terms
Researchers rarely say "cramming." They say massed practice: packing all your study of something into one block. The alternative is distributed or spaced practice, which spreads the same total study time over several sessions with gaps in between.
Cramming is massed practice with a deadline attached. Kornell's 2009 paper defines it as "massing study on the last day before the test." That definition separates two questions people tend to mix up:
- Total time. More hours of study generally beat fewer.
- Timing. For the same number of hours, does it matter whether they come in one block or spread out?
The cramming research is about the second question, and the answer goes back to the start of memory science. Hermann Ebbinghaus, who memorized lists of nonsense syllables for his 1885 book on memory, found that 38 repetitions spread over three days did as much for a 12-syllable list as 68 repetitions packed into one day. We walk through his experiments in our piece on the forgetting curve.
The finding is now called the spacing effect, and it's one of the most replicated results in psychology. A 2006 review by Nicholas Cepeda, Harold Pashler and colleagues gathered 839 assessments of distributed practice from 317 experiments in 184 articles. Spaced study beat massed study across that literature, and the best gap between sessions got longer as the final test got further away.
Is Cramming Effective? What the Studies Found
Three kinds of experiments show the same pattern.
Studies that vary when the test happens. Katherine Rawson and Walter Kintsch had undergraduates read a 1,730-word text about carbon sequestration either once or twice. The second reading came either straight after the first (massed) or a week later (distributed). On a test given right after the last reading, rereading in a row helped compared with reading once, and in their first experiment the massed rereaders also recalled more than the distributed group. On a test two days later, the distributed group came out ahead in both recall and comprehension, and the massed rereaders did no better than students who had read the text once. Hence the paper's title: "Rereading Effects Depend on Time of Test."
A 2019 classroom study by Carla Greving and Tobias Richter found the same shape with 191 seventh-graders. Massed rereading won on an immediate test. A week later the advantage was gone, because the students who spaced their rereading forgot less.
Flashcard studies. In Kornell's experiments, students learned GRE-type word pairs with a web-based flashcard program and took the final test about a day after their last session. Studying one large stack, which puts more time between repetitions of each card, beat studying four smaller stacks separately: 49% recalled versus 36% in the first experiment. In the second, words crammed on the last study day were recalled at 34%, against 54% for words studied across all four days.
Studies that compare rereading with testing. Henry Roediger and Jeffrey Karpicke had students reread one prose passage and practice recalling another. Five minutes later, the reread passage looked better: 81% recalled versus 75%. Two days later the tested passage led 68% to 54%, and after a week, 56% to 42%.
Their second experiment is the one to remember. Some students studied a passage for one five-minute period and then tried to recall it three times. Others spent all four periods studying it. A week later the recall group remembered 61% and the study-only group 40%, even though the study-only group had read the passage 14.2 times on average to the recall group's 3.4. Round after round of rereading the same pages, with no self-testing, is a fair description of most cram sessions.
| Study | Participants and material | Massed or reread | Spaced or tested | Result |
|---|---|---|---|---|
| Ebbinghaus (1885) | Himself, lists of nonsense syllables | 68 repetitions in one day | 38 repetitions over three days | The same learning from 30 fewer repetitions |
| Rawson and Kintsch (2005) | Undergraduates, a 1,730-word science text | Second reading right away | Second reading a week later | Massed beat a single reading on an immediate test; only spaced beat it two days later |
| Roediger and Karpicke (2006) | Undergraduates, prose passages | Four study periods | One study period, then three recall tests | 40% vs. 61% remembered after a week |
| Kornell (2009) | Students, GRE-type word pairs | Small stacks, or cramming on the last day | One large stack, or study spread over days | Crammed words 34% vs. spaced 54% the next day |
| Greving and Richter (2019) | 191 seventh-graders, in class | Massed rereading | Distributed rereading | Massed won immediately; no difference a week later |
These are controlled studies with short materials, not whole courses, and real exams also ask you to apply ideas, which these tests barely measured. But none of them found cramming ahead on a test given a day or more later.
Why Cramming Feels Like It's Working
If spacing works better, why do most learners believe the opposite? Because cramming produces a feeling that's easy to mistake for learning.
Read a chapter three times in a row and the third pass feels smooth. You recognize every sentence. That fluency is real, but you're measuring it minutes after your last look at the page. The exam asks something harder: can you produce this from nothing, days later, without the page in front of you?
Researchers have measured that gap between feeling and result directly:
- Kornell (2009). Across 70 participants, spacing produced better recall for 63. Yet after the first study session, 42 of the 58 people who thought one method had worked better (72%) picked massing.
- Kornell and Robert Bjork (2008). People learned to recognize painters' styles from examples shown either back to back or mixed in with other artists' work. In the first experiment, 78% did better with the mixed presentation, and yet the same share, 78%, said massing had been as good or better. They gave that answer after a test on which mixing had worked better for most of them.
- Roediger and Karpicke (2006). Students who spent all four periods rereading predicted they'd remember the passage better than the students who tested themselves. A week later, they remembered the least.
Robert and Elizabeth Bjork's "new theory of disuse" offers the clearest explanation. It splits memory into two strengths. Retrieval strength is how readily something comes to mind right now; it can shoot up during a session and sink within days. Storage strength is how deeply it's learned; it builds up with study and, in the theory, is never lost.
Cramming drives retrieval strength up fast, which is the confident feeling you have at midnight. It adds far less storage strength, and storage strength is what you're relying on a week later.
The theory's second half explains why spacing feels worse than it is. Pulling something back after you've partly forgotten it adds more storage strength than recalling it while it's still fresh. So the sessions that feel least productive, where you're straining to remember, are often doing the most.
Why Most Students Cram Anyway
In 2012 Marissa Hartwig and John Dunlosky asked 324 undergraduates at Kent State University how they study. A few of the answers:
| Survey item | Answer | Share of students |
|---|---|---|
| Strategies used regularly | "Cram" lots of information the night before the test | 66% |
| Strategies used regularly | Underline or highlight while reading | 72% |
| How do you decide what to study next? | Whatever's due soonest or overdue | 56% |
| How do you decide what to study next? | I plan my study schedule ahead of time | 13% |
| Spacing or massing? | I most often do my studying in one session | 53% |
The deadline answer explains the most. When Kornell and Bjork asked UCLA students the same question in 2007, 59% chose it too. If you always study whatever is due next, you'll cram by default, because the nearest deadline gets all your attention right before it arrives. Hartwig and Dunlosky also found that low performers were more likely to study late at night, and that deadlines drove everyone's studying, low performers' most of all.
So cramming is less a character flaw than what you'd expect from deadline-driven scheduling when nothing else is planned. The fix is a calendar decision, not a willpower one.
The 72% who highlight deserve a second look too. In a 2013 review, Dunlosky and colleagues rated highlighting "low utility," next to rereading; the two "high utility" techniques were practice testing and distributed practice. A highlight is a marker for what to come back to, and it only pays off if you come back.
Should You Pull an All-Nighter Before an Exam?
Usually not. Cramming and the all-nighter are separate decisions: you can cram from 6 p.m. to 11 p.m. and still sleep. The all-nighter adds a second cost on top of massed practice, and the research on that cost is blunt.
Missing sleep before learning. In a 2007 study from Matthew Walker's lab (Yoo and colleagues), volunteers stayed awake for about 35 hours, then viewed images inside an MRI scanner. Two days later, after catching up on sleep, they remembered the images 19% worse on average than volunteers who had slept normally, and Walker said the deficit reached as much as 40%. You'll often see "40%" quoted on its own; it's the high end, not the average.
Missing sleep after learning. In 2000, Robert Stickgold and his colleagues James and Hobson trained people on a visual discrimination task. People kept awake the first night after training showed no significant improvement when tested after two further nights of normal sleep. The catch-up sleep didn't make up for the first night. That was a perceptual skill, but a similar pattern shows up for facts: Gais, Lucas and Born found in 2006 that memory for vocabulary was better when sleep followed within a few hours of learning.
The next day. In a 2013 study, Gillen-O'Neel, Huynh and Fuligni had 535 high school students keep daily diaries for 14 days in each of 9th, 10th and 12th grade. Nights when a student gave up sleep to study more than usual were followed by more trouble understanding what was taught in class, and a higher chance of struggling on a test or assignment, the next day. That held no matter how much the student usually studied.
Your own judgment. Pilcher and Walters gave 44 college students a critical-thinking test after either 24 hours awake or about 8 hours of sleep. The sleep-deprived students scored significantly worse. They also rated their own concentration and effort higher, and estimated their performance as significantly better. The all-nighter makes you worse at the test and worse at noticing it.
Pamela Thacher's 2008 survey of 120 students at St. Lawrence University adds a correlational data point: students who had pulled all-nighters averaged a 2.95 GPA, against 3.20 for those who never had. It can't show cause, but it fits.
We cover what happens to memory overnight in more depth in sleep and learning.
How to Cram for an Exam When You Have To
Sometimes you're simply out of time. This won't turn one night into a semester of studying, but it makes the night count for more.
- Triage first. Spend 15 minutes listing every topic the exam covers, using the syllabus, study guide or past papers. Rank them by how much they're worth and how shaky you are. You won't cover everything, so choose on purpose instead of by page order.
- Swap rereading for retrieval. Read a section once, close it, and write down everything you remember. Then check and fill the gaps. This is the blurting method, and it's the cram-night version of the Roediger and Karpicke result: fewer passes over the material, more attempts to pull it out of your head. Practice questions and past papers do the same job.
- Shuffle, don't block. Keep flashcards in one big stack instead of drilling small piles to perfection, the within-session spacing that worked in Kornell's experiments. For problem sets, mix problem types instead of doing ten of the same kind in a row, much as mixing artists helped people learn painters' styles in Kornell and Bjork's study.
- Split the night into rounds. Two or three passes over the material, with breaks or another subject in between, beat one unbroken block. The gaps will be short, so you'll get less of the spacing benefit. Recall practice still pays off with short gaps (in Roediger and Karpicke's study the three recall tests came minutes apart), but spreading recall rounds further apart helps more.
- Stop, then sleep. Do a last recall round on your weakest topics and go to bed. In a 2016 study by Mazza and colleagues, people learned Swahili words paired with French translations, then relearned them 12 hours later. Those who slept in between needed about half as many rounds to relearn the list (3.05 versus 5.80). Six months later they recalled 8.67 of 16 words, against 3.35 for the group that learned and relearned within one day.
- Recall in the morning. Don't reread. Blurt the main topics from memory for 20 minutes, check what's missing, and leave it there. Recall also holds up better under pressure. In a 2016 Tufts study by Smith, Floerke and Thomas, people who had learned by retrieval practice showed no memory impairment when put under stress a day later. People who had learned by restudying remembered fewer items, especially under stress.
- If you're anxious, write it out. In two classroom studies of ninth-grade biology students, Ramirez and Beilock found that students in the more test-anxious half who spent 10 minutes writing about their worries right before the final averaged a B+, against a B- for equally anxious classmates who didn't write.
Here's how that looks for a 9 a.m. exam:
| Time | What to do | Why |
|---|---|---|
| 6:00 p.m. | List topics, rank by weight and weakness | You can't cover everything, so choose |
| 6:15 to 8:00 p.m. | Round 1: read each topic once, then blurt or answer practice questions | Recall beats rereading (Roediger and Karpicke, 2006) |
| 8:00 to 8:30 p.m. | Break, food, or a different subject | Puts a gap before round 2 |
| 8:30 to 10:30 p.m. | Round 2: recall everything, focus on what you missed | Spread-out repetitions beat back-to-back ones, even within one session (Kornell, 2009) |
| 10:30 to 11:00 p.m. | Weakest topics only, from memory | Sleep soon after studying helps you keep it (Gais and colleagues, 2006) |
| 11:00 p.m. | Sleep at least 7 hours | Sleep between sessions halved relearning (Mazza and colleagues, 2016) |
| 7:30 a.m. | 20 minutes of recall, no rereading | Material learned by recall held up under stress (Smith and colleagues, 2016) |
| Just before the exam | 10 minutes writing about your worries, if anxious | B+ vs. B- for anxious students (Ramirez and Beilock, 2011) |
What you're cutting is the part that feels most like studying: rereading. It's also one of the techniques the research supports least.
Cramming vs. Spaced Repetition: How to Stop Needing to Cram
| Factor | Cramming (massed) | Spaced study |
|---|---|---|
| Test a few minutes later | Often as good, sometimes better | Similar or lower |
| Test the next day | Already behind (34% vs. 54% in Kornell, 2009) | Ahead |
| Test days or weeks later | Falls off quickly | Holds up much better |
| How it feels | Fluent and productive | Harder, less certain |
| What learners predict | Usually rated as more effective | Usually underrated |
| Sleep | Often sacrificed | Protected |
| Cumulative final | Relearn much of it | Review what you kept |
The last row is where cramming costs the most. Midterm material comes back on the final, and in the next course that assumes you know it. Anything you crammed and never reviewed has to be learned again.
How far apart should the sessions be? Cepeda and colleagues tested more than 1,350 people in 2008 and found the best gap depends on when the test is. For a test a week away, the best gap was about one to three days. For a test a year away, it was about three to four weeks. Getting the gap right paid off: for the same study time, the best gap produced 64% more recall, averaged across the four test delays, than a review that started straight after the first session.
You don't need to hit those gaps exactly. Two fixed schedules turn the idea into a plan:
- The 2357 method. Learn a topic on day 1, then review it on days 2, 3, 5 and 7. It works as a weekly cycle during term, and it has a countdown version for a fixed exam date.
- Glasp's free Spaced Repetition Calculator. Enter a start date and your exam date. It lays out reviews at 1, 3, 7, 14, 30, 60 and 120 days, drops any that fall on or after the exam day, and adds a final review the day before. You can download the schedule as a calendar file.
That last review, the day before the exam, keeps the one good idea inside cramming: a fresh look at the material right before the test, without the all-nighter.
The other half is making each review quick enough that you'll do it:
- Highlight with a purpose. Glasp's web highlighter works on web pages and PDFs, so assigned readings and PDF lecture slides end up in one place. When a review day comes, you have a short list of passages that mattered instead of 300 pages to reread. Turn each highlight into a question and answer it before you look.
- Shrink long lectures. For lectures and explainers on YouTube, YouTube Summary gives you a summary and the full transcript, which you can read and highlight. Reviewing the key points takes minutes instead of rewatching the whole video.
- Check yourself against your sources. Glasp's AI chat searches everything you've highlighted to answer a question and points you to the sources it drew on. Answer a question from memory first, then ask the chat the same thing, compare, and open those sources to see what you missed.
Frequently Asked Questions
Is cramming better than not studying at all?
Yes, if the test is close. Massed rereading beat a single reading on an immediate test in Rawson and Kintsch's study, and Kornell's crammed words were still recalled at 34% the next day. But two days later, the massed rereaders did no better than students who had read the text once, while students who spaced their reading were ahead of both. A cram beats doing nothing; studying earlier beats a cram.
Is cramming bad for your brain?
We found no study showing that a single cram session harms your brain. The costs are practical: crammed material fades within days, and cramming often eats into the sleep your brain uses to stabilize what you studied. If you cram and still sleep, the main loss is retention.
Is it better to sleep or study the night before an exam?
Sleep, for most students. In a 14-day diary study of 535 high school students, nights of extra studying bought with lost sleep were followed by more academic problems the next day. Sleep-deprived college students also scored worse on a critical-thinking test while rating their performance higher. Study until a reasonable hour, do one final recall round, and go to bed.
How long does crammed information last?
Not long. In Rawson and Kintsch's study, the advantage of massed rereading had disappeared two days later. In Roediger and Karpicke's experiment, recall for students who only reread was 52% lower after a week than after five minutes, compared with 14% lower for students who practiced recall. Expect to lose about half of what you crammed by rereading within a week unless you review it.
Can you cram in one night and still pass?
Often, yes, which is why the habit survives. A night of cramming beats not studying, even if the same hours spread over a week would have done better. Your odds improve if you test yourself instead of rereading and still sleep. The risk comes after the exam: cumulative finals and later courses that assume you still know the material.
Should you study the morning of an exam?
Briefly, and by recalling rather than rereading. Spend 20 minutes writing out the main topics from memory and checking the gaps. In a 2016 Tufts study, memory built by retrieval practice held up under stress better than memory built by restudying, so recall is the better warm-up.
Why Cramming Survives, and How to Need It Less
Cramming survives because it beats doing nothing on the one test that comes right away. It feels productive, it often gets you a passing grade, and the cost doesn't show up until later, on the final and in the next course.
Move some of the hours you'd spend cramming to earlier days, and turn every session into retrieval instead of rereading. Same hours, different calendar. When you can't do that, cram the smart way: test yourself, mix it up, and protect your sleep.
Glasp makes studying early easier. Highlight readings and YouTube lectures as the term goes on, so review days start from a short list of what mattered. Then plan the reviews with the Spaced Repetition Calculator, and check your answers against your own sources with Glasp's AI chat.
Cram if you must. Then sleep. Then start the next unit a week earlier.
