How to Memorize Formulas for Math, Physics, and Chemistry

The best way to memorize formulas is to need fewer of them. Learn a small set of core formulas well enough to derive the rest, write them from memory on a blank page instead of rereading a formula sheet, and use them in mixed practice problems on a spaced schedule. Flashcards help too, as long as the front describes a situation rather than the formula's name.

Most students do the opposite. They copy a formula sheet, read it over the night before, and then freeze when a problem doesn't say which formula it wants. Recognizing a formula on a sheet and producing the right one for a problem are different skills.

This guide shows how to memorize math, physics, and chemistry formulas in five steps, with derivation chains, unit checks, example cards, and a seven-day plan.

Why formulas slip away during exams

Three things go wrong, and each one has its own fix.

1. You memorized a string of symbols. Without meaning attached, s = ut + ½at² and s = ut + ½at look equally plausible under pressure. Knowing what each term means, and what units it has, tells you which one is real.
2. You practiced recognition, not recall. Reading a formula sheet makes every formula look familiar. Familiar isn't the same as retrievable from a blank page, and the exam gives you a blank page. Our guide to active recall explains why testing yourself beats rereading.
3. You never practiced choosing. Homework sets often group problems by formula. On the exam, picking the formula is half the problem.

Step 1: Derive, don't memorize

Most formula lists are a few core relationships plus their consequences. If you can rebuild the consequences, you only need to store the core, and you have a backup when memory fails.

Here's the classic example from physics: the constant-acceleration (SUVAT) equations, where u is initial velocity, v is final velocity, a is acceleration, t is time, and s is displacement.

1. Start from the definition of acceleration. a = (v − u) / t, so v = u + at.
2. Displacement is average velocity times time. With constant acceleration, the average velocity is (u + v) / 2, so s = ½(u + v)t.
3. Substitute step 1 into step 2. s = ½(u + u + at)t, which gives s = ut + ½at².
4. Eliminate time. From step 1, t = (v − u) / a. Put that into step 2: s = (v + u)(v − u) / (2a) = (v² − u²) / (2a), so v² = u² + 2as.

Two ideas you already understand produce four equations. Practice the derivation until you can do it in under a minute, and you'll never need to memorize the last two cold. The same trick works across subjects:

Deriving is slower than recalling, so you don't want to derive everything during an exam. The point is to derive each formula several times while studying. That repeated effort makes the formula itself easier to recall, and the derivation is there if recall fails.

Step 2: Learn what every symbol means, with units

A formula is only usable if you know what each letter stands for, its units, and when the formula applies. Write all three next to it.

• v² = u² + 2as: velocities in m/s, acceleration in m/s², displacement in m. Only valid for constant acceleration.
• PV = nRT: temperature in kelvin, always. R = 8.314 J/(mol·K) if P is in pascals and V in cubic meters; R = 0.08206 L·atm/(mol·K) if P is in atmospheres and V in liters. Valid for ideal gases.
• T = 2π√(L/g) for a pendulum: L in m, g in m/s², T in seconds. Valid for small swings.

Then use units as a lie detector. Check s = ut + ½at²: (m/s)(s) gives meters and (m/s²)(s²) gives meters. Both terms match the left side, so the formula is at least the right shape. Now check the wrong version, s = ut + ½at: (m/s²)(s) gives m/s, which can't be added to a distance. A unit check takes seconds and catches this kind of slip every time.

Add a limiting-case check for the formulas you use most. If a = 0, v² = u² + 2as says v = u, which is right: no acceleration, no change in speed. For the pendulum, a longer string gives a longer period, and stronger gravity gives a shorter one. If your remembered version predicts the opposite, it's wrong.

Step 3: Write your formula sheet from memory

Don't copy a formula sheet. Generate it.

1. Take a blank page and write every formula for the chapter that you can remember, with units and conditions. Set a timer for 10 minutes.
2. Open your notes and compare. Mark what you missed or got wrong in a different color.
3. The marked formulas become your study list for the next few days. The rest only need occasional review.
4. Repeat the blank-page sheet every few days until the misses stop.

This matters even if your exam allows a formula sheet or crib card. In one classroom experiment with 51 students, people who expected to use their crib card scored significantly lower on a surprise pretest without it than on the same questions with it, which the authors read as the card reducing what they learned. When students made a card expecting not to use it, they scored the same with or without it. An earlier study of 54 students found that being allowed crib cards didn't improve exam scores at all, even on factual questions.

So build the sheet as if you won't have it. If the exam allows it, it becomes a backup instead of a crutch. If your exam supplies an official reference instead, like the FE exam's searchable handbook, our engineering study tips cover how to practice with it.

Step 4: Make formula flashcards that start from a situation

A card that says "Front: SUVAT equation without time. Back: v² = u² + 2as" trains you to recall the formula when someone names it. Exams don't name it. Put the situation on the front instead:

Keep each card to one decision. Review them with spaced repetition: the same day, the next day, after a few days, then a week, then a few weeks. Our flashcard guide covers card-writing rules, and the spaced repetition guide explains the timing. The same task-first card works for code; see flashcards for remembering coding syntax.

If you'd rather not write cards by hand, an AI flashcard generator can draft them from a chapter PDF or your notes. Rewrite the fronts as situations and check every formula before you trust a generated card.

Constants: the one place to memorize directly

Constants can't be derived, so they belong on plain flashcards. Learn them to the precision your course uses, with units. These values are rounded from NIST's list of constants (CODATA 2022); standard gravity is a defined value, and R in L·atm is a unit conversion:

Mnemonics earn their place here too, for pairings you can't reason out. Nothing about the word "sine" says "opposite over hypotenuse," so SOH-CAH-TOA (sine = opposite/hypotenuse, cosine = adjacent/hypotenuse, tangent = opposite/adjacent) supplies the link. See mnemonics and memory aids for when they help and when they don't.

Step 5: Use formulas in mixed problems

Recall on a flashcard is the warm-up. The real test is pulling the right formula out mid-problem, so most of your formula practice should be problems.

Mix the problem types instead of doing ten of the same kind. In an eight-week study in a university physics course, students whose homework alternated topics scored higher on surprise tests weeks later, and part of the gain was in recalling the equations each problem needed. They also rated the mixed homework as harder. The full results are in our guide to studying physics and biology, and the general method is in the interleaving guide.

When you practice:

• Write the formula before any numbers. Symbolic first, then substitute. You'll spot the wrong formula before you waste time on arithmetic.
• Say why that formula fits. "No time given, so v² = u² + 2as."
• Check units and size on the answer. A stopping distance of 0.06 m or a gas volume of 40,000 L should make you stop and look.

For more on practice sets in quantitative courses, see how to study math effectively.

How to memorize formulas in a week: a 7-day plan

About 30 to 40 minutes a day. Start a week before you need the chapter cold.

After the week, do a blank-page sheet again after about two weeks and again after a month. Each one takes ten minutes and tells you exactly which formulas are fading. Spacing the reviews matters: a review of 317 experiments found that spreading study across sessions beats massing the same amount of study into one.