Does cramming actually work? What science says about last-minute memorizing

In short

  • Cramming can help in the short term, but it is often worse for long-term learning.
  • In a classic study, rereaders did better after 5 minutes (83% vs 71%), but testing yourself won after a week (61% vs 40%).
  • Spreading study over several days beats massing it, and cutting sleep hurts memory formation.
  • If the exam is tomorrow: prioritize, quiz yourself, review what you got wrong, and sleep.

The night before an exam, we feel like we can memorize far more than usual.

Studying that normally takes hours suddenly becomes laser-focused when the deadline is close. You may even have gotten good grades thanks to cramming.

So is cramming actually an effective way to study?

In short, cramming can help with short-term exam preparation, but it is often a disadvantage for long-term learning.

Let's look at why, through learning science research.

1. Why cramming feels like it works

In 2006, researchers at Washington University had college students study science passages, then compared a group that reread the text with a group that repeatedly tested themselves by recalling it.

Memory after 5 minutes vs. after 1 week
Repeated reading5 minutes later: 83%
1 week later: 40%
Feels better now, fades fast
Repeated recall5 minutes later: 71%
1 week later: 61%
Lasts
Source: Roediger & Karpicke, Psychological Science (2006), Experiment 2.

Five minutes later, the rereading group remembered more. A week later, the result flipped completely. The rereaders dropped to 40%, while the testing group held 61%, a gap of 21 percentage points.

This study did not directly compare cramming itself. But it shows something important:

remembering well right now and remembering for a long time are different problems.

Original study

2. For the same time, spreading it out works better

A meta-analysis published in Educational Psychology Review in 2025 examined how the spacing of repeated study affects math learning.

Across 27 studies, spreading study over time produced better results than massing it in one sitting, with an effect size of Hedges' g = 0.28.

That does not mean grades rose by 28%. It is a standardized statistical measure of the difference between the two approaches.

A more recent meta-analysis from October 2026, based on real school classrooms, found an average effect for spaced practice of g = 0.20. So the benefit of spreading review over several days shows up not only in labs but in real classrooms.

One reason is that recalling material after some time has passed helps your memory retain it longer.

2025 math meta-analysis · 2026 classroom meta-analysis

3. The riskiest part of cramming is the all-nighter

When cramming, people often cut their sleep.

But instead of buying study time, losing sleep can reduce your ability to remember new information.

A 2024 meta-analysis pooled 1,234 participants from 39 study reports. When sleep was restricted to 3 to 6.5 hours, the ability to form memories dropped significantly compared with normal sleep, with an effect size of Hedges' g = 0.29.

Sleep loss does not only make you tired. It also affects how you learn new information and stabilize memories. That is why giving up sleep for one more hour of memorizing the night before an exam is not necessarily a good trade.

Neuroscience & Biobehavioral Reviews (2024)

4. If the exam is tomorrow, how should you study?

Of course, if you only have a day left, you cannot do much spacing. That does not mean you should give up on cramming. It means using the remaining time more effectively.

  1. Start with the most important concepts. Rather than reading everything equally, set priorities from the exam scope and past papers.
  2. Close the book and try to recall. Instead of rereading, explain things yourself or work problems. That is retrieval practice.
  3. Review what you got wrong. Spend your time on concepts you cannot explain or keep missing, not on what you already know.
  4. Protect your sleep. Sitting the exam well rested may help your memory and attention more than another hour of cramming.

The most important thing is knowing exactly what you do not know.

You have probably felt it: you understood the material while reading, but when you closed the book and tried to explain it, you went blank. Feeling that you understand and being able to recall and use it are different things.

Conclusion: studying for the exam and studying to keep it are different

Cramming can be a realistic strategy when time is limited. But what you hold in your head right before the exam does not necessarily stay as long-term knowledge.

What matters is not simply how long you studied. Check whether you can recall it yourself, explain it to someone else, and still remember it after time has passed.

Instead of trying to memorize everything perfectly in one go, start finding the parts you do not understand and checking them again and again.

That is the approach behind obbler, an AI study app. You talk with an AI about your own study materials, and it checks your real understanding through quizzes and repeated review.

Check your understanding with your own material →

Study your own material with obbler

Add a PDF, photos, a link or a video. Chat about it and turn it into notes, flashcards and quizzes.

Download on the App Store