# How to multiply two-digit numbers in your head

Canonical: https://romeoapps.app/en/numvolt/multiply-two-digit-numbers-in-your-head/
Author: Roméo Gambino · Published: 2026-09-10 · Updated: 2026-09-10 · App: Numvolt (https://romeoapps.app/en/numvolt/)

Four ways to multiply two-digit numbers without paper: splitting, rounding and correcting, the difference of squares and numbers near 100, with worked examples.

## Why two-digit products feel hard

A single-digit product such as 7 × 8 is usually retrieved from memory rather than calculated. Two-digit products almost never are, so you have to compute, and computing in the head means keeping partial results while producing the next one. DeStefano and LeFevre’s review describes working memory as the bottleneck of mental arithmetic, and multi-digit multiplication loads it heavily. Each method below reduces how much you hold at once. None of them is best for every pair; the skill is choosing the one that fits the numbers in front of you.


## Start with one number times a single digit

Before two digits by two digits, make single-digit multipliers automatic. 47 × 6 is 40 × 6 plus 7 × 6, so 240 + 42 = 282. Working from the tens first, left to right, gives you the large part of the answer early, which is also the part that matters most if you are interrupted. This single step is the building block of every method that follows.


## Method one: split one number

The general method is to split one factor into tens and units and multiply each part. For 34 × 27, split 27 into 20 and 7: 34 × 20 = 680 and 34 × 7 = 238. Then 680 + 238 = 918. You only ever hold one running total and one new partial product, which is why this method works for any pair. Split whichever number gives the easier partial products; splitting 34 into 30 and 4 would also work, giving 810 + 108 = 918.


## Estimate before you calculate

Before starting, round both numbers: 34 × 27 is close to 30 × 30, so the answer should be near 900. That takes a second and catches the worst errors, such as a result of 9,180 or 91.8 caused by a slipped place value. It also gives your attention a target, so the final addition feels like a confirmation rather than a leap.


## Method two: round and correct

When one factor is close to a round number, multiply by the round number and then correct. 29 × 34 is 30 × 34 minus one 34: 1,020 − 34 = 986. 48 × 15 is 50 × 15 minus 2 × 15: 750 − 30 = 720. The correction is usually a small multiplication, and the round product is quick. The method saves the most when the factor sits one or two away from a multiple of ten.


## When rounding is the wrong move

Round and correct loses its advantage when the correction itself is large. 36 × 47 rounded to 36 × 50 needs a correction of 36 × 3 = 108, giving 1,800 − 108 = 1,692. That works, but splitting, 36 × 40 + 36 × 7 = 1,440 + 252 = 1,692, is no slower and involves no subtraction. As a rule of thumb, round when the distance to the round number is one or two, and split otherwise.


## Method three: the difference of squares

When two numbers sit an equal distance either side of a round number, their product is the square of the middle number minus the square of the distance. This comes from the identity (a − b) × (a + b) = a² − b². 48 × 52 is 50² − 2² = 2,500 − 4 = 2,496. 37 × 43 is 40² − 3² = 1,600 − 9 = 1,591. 46 × 54 is 2,500 − 16 = 2,484. The method only needs the squares of round numbers, which are easy, and the squares from 1 to 9, which you already know.


## When the pair is not symmetric

If the two numbers are not the same distance from a round middle, the idea still works with a small correction. 47 × 52 is 47 × 53 minus 47, and 47 × 53 is 50² − 3² = 2,491, so the answer is 2,491 − 47 = 2,444. Whether this beats splitting depends on the person. Try both on the same problem and keep the one that feels lighter.


## Method four: numbers just below 100

For two numbers close to 100, work with how far each is below 100. For 96 × 97, the distances are 4 and 3. Subtract one distance from the other number: 96 − 3 = 93, which gives the first two digits. Multiply the distances: 4 × 3 = 12, which gives the last two digits. The answer is 9,312. The same pattern works above 100: for 103 × 107, add 103 + 7 = 110 and multiply 3 × 7 = 21, giving 11,021. If the product of the distances has only one digit, write it with a leading zero: 98 × 97 gives 95 and 06, so 9,506.


## Why the near-100 method works

It is the same algebra as splitting, arranged to exploit the round number: (100 − a) × (100 − b) = 100 × (100 − a − b) + a × b. The first part counts the hundreds and the second part is what remains. Understanding why it works matters more than memorising it. A method you understand can be adapted, for instance to numbers near 50 or near 1,000, while a memorised trick breaks the first time the numbers change.


## The cross method

Another general method works column by column. For 43 × 21, the hundreds are 4 × 2 = 8, the tens are 4 × 1 + 3 × 2 = 10, and the units are 3 × 1 = 3. Combine them: 800 + 100 + 3 = 903. It is the written algorithm performed in your head, and it suits people who like a fixed routine. Its weakness appears when the middle total is large and spills into the hundreds, because you must hold it while combining, which is exactly the load the other methods avoid.


## Choosing a method in two seconds

Look at the pair before you start. An equal distance from a round number: difference of squares. One number next to a multiple of ten: round and correct. Both close to 100: the near-100 method. Anything else: split the easier number. This choice is itself a skill, and it develops best on mixed problems rather than twenty of the same kind. Rohrer and Taylor found that shuffling problem types during practice improved later test performance compared with practising one type at a time.


## Check the answer cheaply

Two checks take a second each. The last digit: 34 × 27 must end in 8, because 4 × 7 = 28. The size: the estimate of 900 says that 918 is plausible. Neither check proves the answer right, but together they catch most slips, and they are fast enough to use every single time.


## Accuracy first, then speed

Time pressure pushes people back to their slowest, most familiar method. Work without a timer until choosing a method is automatic, then add a clock. When you get a problem wrong, redo that same problem correctly straight away rather than moving on, so that the last version you rehearse is the right one.


## Where Numvolt fits

Numvolt’s multiplication sessions give you problems to practise the choice of method on. In Zen mode there is no timer, and a wrong answer shows the method behind the correction, which is the moment to compare your route with another one. Timed sprints of 15, 30 or 60 seconds then show whether the method holds under a clock, and a duel on one iPhone turns the same problems into a game with someone else.


## Questions

### What is the fastest way to multiply two-digit numbers in your head?

There is no single fastest method. Split one number into tens and units by default, and switch to rounding and correcting, the difference of squares or the near-100 method when the numbers suit them.

### How do you multiply 48 × 52 in your head?

Both numbers are 2 away from 50, so the product is 50² − 2² = 2,500 − 4 = 2,496.

### How can I check a mental multiplication?

Compare the last digit with the product of the two last digits, and compare the size with a rounded estimate. For 34 × 27, the answer must end in 8 and be near 900.

### Should I practise with a timer?

Only once the methods are reliable. Untimed practice builds the method; a timer then measures whether it has become quick.

## Further reading

- DeStefano & LeFevre (2004), The role of working memory in mental arithmetic, European Journal of Cognitive Psychology — https://doi.org/10.1080/09541440244000328
- Rohrer & Taylor (2007), The shuffling of mathematics problems improves learning, Instructional Science — https://doi.org/10.1007/s11251-007-9015-8