In this article
Picture a student preparing for a math exam. They have problems on volumes of spheres, cones, cylinders, and prisms. The intuitive plan is to do all the sphere problems first, then all the cone problems, then cylinders, then prisms — one topic mastered at a time. By the end of the sphere block, every answer comes easily. It feels like learning.
That approach is called blocked practice, and decades of cognitive science suggest it produces a comforting illusion of competence that rarely survives to test day. The alternative — interleaving, or deliberately mixing problem types together — feels clumsier and slower in the moment, yet consistently produces stronger, more durable learning.
This article explains the difference, the mechanism behind it, the key studies that established it, and how you can restructure your own practice to take advantage of it.
Two Ways to Structure Practice
Blocked vs Interleaved
The distinction is purely about order, not content. Imagine three topics — call them A, B, and C. In blocked practice you finish all of one before touching the next: A A A B B B C C C. In interleaved practice you cycle through them: A B C A B C A B C. Same problems, same total volume of work — only the sequence changes.
- –One topic mastered before the next
- –Feels smooth and efficient in the moment
- –Strategy is obvious — no need to choose
- –Weaker retention and transfer to the test
- +Topics mixed throughout the session
- +Feels harder and slower — more errors
- +Forces you to identify which strategy applies
- +Stronger retention and transfer to the test
Notice the central paradox already taking shape: the version that feels worse while you do it is the one that works better when it counts. That tension is the heart of this entire topic, and it has a name — a desirable difficulty.
The Mechanism
Discrimination and Strategy Selection
Why would simply shuffling the order of problems matter so much? The leading explanation is that interleaving trains a skill that blocked practice quietly skips: choosing the right approach.
When you work a block of sphere problems, you never have to ask “what kind of problem is this?” You already know — they're all spheres. You just reach for the formula you used on the previous problem and apply it again. Your brain practices executing a procedure, but it never practices the step that comes first on a real exam: discriminating between problem types and selecting the correct strategy from many competing options.
Interleaving puts that discrimination step front and center. Because the next problem could be anything, you have to inspect each one, notice the features that distinguish a cone from a cylinder, and retrieve the matching method. Robert Bjork and colleagues describe this as one of several desirable difficulties: conditions that slow learning down in the short term but strengthen it in the long term, precisely because the extra effort engages deeper processing. Interleaving also bundles in two other proven effects — it naturally spaces your exposure to each topic over time, and it forces repeated retrieval of the right strategy rather than mechanical repetition.
“Conditions of learning that make performance improve rapidly often fail to support long-term retention and transfer, whereas conditions that introduce difficulties for the learner — though appearing to slow the rate of learning — often optimize it.”— Robert A. Bjork, on desirable difficulties
The Evidence
Math, Art, and Beyond
The interleaving effect isn't a single curiosity — it has been replicated across very different kinds of material. Two studies are especially instructive.
Rohrer & Taylor (2007): Mixing Math Problems
Doug Rohrer and Kelli Taylor taught students to calculate the volumes of four obscure geometric solids. One group practiced in blocks (all of one solid, then all of the next); another practiced the identical problems interleaved. During practice, the blocked group performed far better — they looked like they were learning faster.
On a test one week later, the result flipped dramatically. The interleaved group answered roughly 63% of problems correctly; the blocked group managed only about 20%. The very practice schedule that looked inferior during training produced more than triple the test performance.
Kornell & Bjork (2008): Learning Painters' Styles
Nate Kornell and Robert Bjork tested whether the effect held for inductive learning rather than procedures. Participants studied paintings to learn to recognize the styles of different artists. Some saw an artist's works grouped together (blocked); others saw works by different artists interleaved.
When later shown new paintings and asked to name the artist, the interleaved learners were significantly more accurate at classifying styles they had never seen before. Tellingly, most participants believed blocking had helped them more — their judgment ran exactly opposite to their actual performance.
The pattern recurs across mathematics, category learning, motor skills, and beyond. Interleaving tends to depress performance during practice and improve it on a delayed test — the opposite of what blocked practice does. Which raises the obvious question: if it works so well, why does almost no one do it naturally?
“The conditions that make practice feel productive are often the very ones that make it forgettable.”
Why It Feels Worse But Works Better
The Fluency Illusion
The answer is a cognitive trap known as the fluency illusion. When you practice in a block, each problem flows easily because you've just done a nearly identical one. That smoothness feels like mastery. But fluency during practice is a poor predictor of retention — it measures how easy the material is right now, with all the context handed to you, not how well you'll reconstruct it later when the cues are gone.
Interleaving strips away that artificial smoothness. You make more mistakes, you pause to think, you occasionally pick the wrong method and have to correct course. It feels like you're learning less. In reality, that friction is the learning — each act of discrimination and retrieval is laying down a more durable, more flexible memory. This is why students in the Kornell & Bjork study confidently rated the worse method as better: they were reading their own fluency instead of their actual knowledge.
The practical danger is that the fluency illusion makes blocked practice self-reinforcing. It feels good, so we keep choosing it — and we mistake the comfort for progress right up until the exam reveals the gap.
When Blocked Practice Still Helps
Don't Throw It Out Entirely
Interleaving is not a blanket replacement for blocking. The evidence points to a sensible sequence: blocked practice is genuinely useful during initial skill acquisition, when you are first learning what a procedure even is.
Before you can choose between strategies, you have to possess them. If you've never computed the volume of a cone, interleaving cone problems with five other types from the very first minute just produces confusion — there's no stored procedure to retrieve yet. A short block to establish the basic method first, then a switch to interleaving for consolidation, captures the best of both. Block to build the skill; interleave to sharpen and retain it.
A simple rule of thumb
Use a brief block when a concept is brand new and you're still grasping the basic mechanics. As soon as you can do a couple in a row without help, start mixing it back in with everything else. The goal is to spend most of your practice in the interleaved zone, not the blocked one.
How to Interleave Your Own Study
Problem Sets, Subjects, and Review
Putting interleaving into practice is mostly a matter of resequencing work you were going to do anyway. Three concrete tactics cover most situations.
Mix problem types within a subject
Instead of working a textbook chapter as a solid block, pull a handful of problems from several chapters you've already started and shuffle them. Solve one of each type in rotation. The moment of asking "which method does this need?" is the part that's building the skill that matters on the exam.
Alternate subjects in a study session
Rather than spending a three-hour block on chemistry and the next on statistics, break the time into shorter rotating segments — chemistry, then statistics, then back again. This spaces each subject across the session and keeps the strategy-selection muscle engaged across very different material.
Interleave your review and self-testing
When reviewing flashcards or doing practice questions, resist the urge to drill one deck or one topic to exhaustion. Shuffle older material back in with the new. Cumulative, mixed review forces repeated retrieval and prevents the false fluency that comes from seeing the same topic over and over.
A few cautions make the difference between productive interleaving and pointless churn. Interleave related material — topics that could plausibly be confused with one another — rather than wildly unrelated subjects, since the benefit comes from learning to tell them apart. And expect it to feel harder: if your practice feels effortless, you may have drifted back into a block. The difficulty is a feature, not a bug.
The Bottom Line
Blocked practice optimizes for how good you feel during a study session. Interleaving optimizes for how much you actually know a week, a month, or a semester later. Rohrer and Taylor's math students and Kornell and Bjork's art learners tell the same story: mixing topics depresses practice performance and improves real, transferable learning.
The catch is that interleaving requires trusting the evidence over your own in-the-moment sense of progress. The fluency illusion will keep whispering that blocking is working better. It isn't. Build a new skill with a short block, then mix it relentlessly with everything else — and let the discomfort be your signal that the learning is sticking.
References
Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In Metacognition: Knowing About Knowing (pp. 185–205). MIT Press.
Bjork, E. L., & Bjork, R. A. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. In Psychology and the Real World (pp. 56–64). Worth Publishers.
Kornell, N., & Bjork, R. A. (2008). Learning concepts and categories: Is spacing the “enemy of induction”? Psychological Science, 19(6), 585–592.
Rohrer, D., & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science, 35(6), 481–498.
Rohrer, D. (2012). Interleaving helps students distinguish among similar concepts. Educational Psychology Review, 24(3), 355–367.
Turn your lectures into interleaved practice automatically
Interactive Lectures builds mixed, spaced practice sets from your lecture content — so your review naturally interleaves topics and forces the retrieval that makes learning stick, without you having to reshuffle anything by hand.
Request Early Access