Long multiplication is the column method: break the multiplier into its digits, multiply the whole multiplicand by each one, shift each result one place further left than the last, and add the rows together. This long multiplication calculator does not just hand back the product. It prints the complete grid — top number, bottom number, every partial product in its correct column, the rule-off line and the total — the way it would appear in an exercise book, and narrates the digit-by-digit multiplications with their carries so a wrong answer can be traced to the exact digit that caused it.
Arb Digital built this as the multiplication counterpart to our long division calculator, which prints the bring-down, multiply and subtract blocks of the division layout. Both pages exist for the same reason: the answer alone teaches nothing when your own working disagrees with it, and a marker who wants to see method will not accept a bare number. The layout is the point.
What This Long Multiplication Calculator Does
It accepts two numbers, positive or negative, whole or decimal, and produces the standard vertical layout. Internally the numbers are stripped of their sign and decimal point, multiplied as digit strings so that very large values never lose precision to floating-point rounding, and the decimal point is placed back at the end using the simple rule that the product has as many decimal places as the two inputs put together. The sign is applied last.
Alongside the grid you get four supporting figures: how many partial product rows the layout contains, how many digits the answer has, how many single-digit multiplications were performed in total, and a casting-out-nines verification. That last one is a genuine independent check rather than a restatement of the same calculation, and it is explained further down the page.
This page multiplies two numbers in the column layout. It is not the same job as our cross multiplication calculator, which solves a proportion by multiplying across a pair of fractions, and it is not the same as our polynomial division calculator, which works with algebraic terms rather than digits.
How to Use It
- Type the larger number on top. The method works either way round, but putting the number with fewer digits underneath produces fewer partial product rows and less to add up.
- Type the multiplier underneath. Every digit of it generates one row, so a three-digit multiplier means a three-row layout.
- Choose your placeholder style. Writing the zeros makes the leftward shift visible; leaving the columns blank matches how many older textbooks set the work out.
- Read the narration first if your answer differs. It lists each single-digit multiplication with its carry, and almost every hand error is a mislaid carry rather than a mistaken product.
- Confirm with the nines check. If the digit sums do not agree, there is an arithmetic slip somewhere in the layout, though the check cannot say where.
The Method and How It's Calculated
The method rests on distributing a multiplication over addition. Writing 46 as 40 + 6 means that 237 × 46 is 237 × 6 plus 237 × 40. Those two products are the two rows of the layout. The first is 1,422. The second is 9,480, which is 237 × 4 = 948 shifted one column left — that shift is the whole reason the zero placeholder appears. Adding 1,422 and 9,480 gives 10,902, which is the product.
Each row is itself built digit by digit, right to left, carrying into the next column. For 237 × 6: 7 × 6 = 42, so write 2 and carry 4; 3 × 6 = 18, plus the carried 4 makes 22, so write 2 and carry 2; 2 × 6 = 12, plus the carried 2 makes 14, written in full because there is no further column. Reading the written digits from the last one back gives 1422. The narration in the results panel prints exactly this sequence for every row.
Decimals need no new method at all. Multiply as though the points were not there, then count the decimal places in both inputs and put that many into the answer. 2.37 × 4.6 uses the identical grid as 237 × 46; the product 10902 gains three decimal places, giving 10.902. This is why the calculator scales the inputs to whole numbers before it starts, and it is the same reasoning that underpins the place-value arguments in the MIT OpenCourseWare course on Principles of Discrete Applied Mathematics.
Casting Out Nines: A Real Check, Not a Ritual
Add the digits of a number repeatedly until one digit remains, and you have its digital root, which is the remainder on division by nine with nine standing in for zero. The useful fact is that digital roots survive multiplication: the digital root of a product equals the digital root of the product of the two digital roots. For 237 × 46, the digital roots are 3 and 1, whose product is 3; the answer 10,902 has digit sum 12, whose digital root is 3. They match, so the answer passes.
What the check catches is any error that changes the digit sum — a wrong product, a dropped carry, a misaligned row. What it cannot catch is an error that leaves the digit sum unchanged, most commonly two transposed digits, since swapping 10,902 for 10,092 alters nothing about the sum. So a pass is evidence and not proof. It costs a few seconds and it eliminates most careless slips, which is a good trade in an exam.
The same modular reasoning underlies far more serious machinery. Remainder arithmetic is the basis of the checksum digits on bank account numbers and of the modular routines behind our modular exponentiation calculator and modulo calculator. Casting out nines is simply the friendliest instance of it, working in base ten because ten leaves a remainder of one when divided by nine.
Why the Zero Placeholder Matters More Than It Looks
The single most common failure in column multiplication is not a wrong times-table fact. It is a row that has not been shifted, so a partial product that should represent tens or hundreds is added as though it were units. The answer that results is wrong by a large and unobvious margin, and because every individual multiplication in it was correct, checking the digit products finds nothing.
Writing the placeholder zeros defends against this because the zero is a visible reminder that this row was produced by a tens digit, not a units digit. The second row of a two-digit multiplication ends in one zero, the third row in two zeros, and so on. If you set the option above to leave the columns blank, notice that the grid still starts the row one column further left; the shift has not gone away, it has merely stopped announcing itself.
A quick sanity check catches most shift errors before you finish. Round both numbers to one significant figure and multiply those: 237 × 46 is roughly 200 × 50 = 10,000, so an answer near 10,900 is plausible and an answer near 2,300 or 109,000 is not. Estimation is not a substitute for the calculation, but it puts a boundary around it.
The Lattice and Grid Alternatives
Column multiplication is one presentation of the same arithmetic, not the only one. The lattice method draws a box grid, writes one digit product in each cell split by a diagonal, and adds along the diagonals; the carries move along the diagonals instead of along a row. Partial products stay separated, so it suits learners who lose track of carries, at the cost of a grid that takes longer to draw.
The area or box method splits each number by place value — 237 becomes 200 + 30 + 7 and 46 becomes 40 + 6 — and fills a rectangle with the six resulting products, which are then summed. It is slower but it shows the distributive law directly, and it generalises without change to multiplying algebraic brackets, which is exactly why it survives into secondary school. Every one of these methods computes the same set of single-digit products; they differ only in how those products are arranged before addition.
None of them is asymptotically fast. Multiplying two n-digit numbers by the column method costs on the order of n² digit products, which is why the answer grid grows so quickly when both numbers are long. Algorithms that beat it — Karatsuba's divide-and-conquer method and its successors — are treated in the MIT OpenCourseWare Introduction to Algorithms course. They only pay off for numbers far longer than anything written by hand, which is why the schoolbook method has never been displaced on paper.
Negative Numbers, Zeros and Other Edge Cases
Signs are handled entirely separately from the digits. Multiply the magnitudes in the usual grid, then apply one rule: two like signs give a positive product, two unlike signs give a negative one. Trying to carry the minus sign through the columns is a reliable route to error, and the layout has no column for it.
A zero digit in the multiplier produces a row of zeros. It is tempting to skip it, and many people do, but the row still consumes a column position — omitting the row entirely without also shifting the next row twice is another version of the placeholder mistake. This calculator prints the zero row so the alignment stays visible, and counts it in the partial product total.
Trailing zeros can be lifted out to save effort. 2,300 × 4,600 is 23 × 46 followed by four zeros, which turns an eight-row problem into a two-row one. The same trick applies to any factor of ten, and it is the arithmetic behind the shift itself. When a factor is a power of ten exactly, no grid is needed at all.
Where the Column Method Still Earns Its Place
Nobody multiplies six-digit numbers by hand for a living. The method survives in education because it makes place value concrete: the shift is not a rule to memorise but a visible consequence of what the digits mean. Learners who understand why the second row moves left rarely misplace a decimal point later, and those who memorised the shift as a ritual frequently do.
It also remains the fastest reliable way to check a machine answer that looks suspicious. Any time a spreadsheet or a script returns a product that seems too large or too small by a factor of ten, the column layout settles it in under a minute and tells you where the error entered. That habit transfers directly to commercial work, where a misplaced decimal in a budget line is expensive and a rounded mental estimate is the thing that catches it. If you are checking figures that feed into campaign spend, our percentage calculator handles the proportional side of the same arithmetic.
Arb Digital builds reporting and forecasting that shows its working, so a figure you cannot explain never reaches a client deck.
Browse All Free Tools Talk To Our TeamCommon Mistakes to Avoid
- Forgetting to shift a partial product — every row after the first moves one further column left, and a missing shift produces an answer that is wrong by hundreds or thousands with no visible arithmetic error.
- Adding a carry before multiplying — the carry is added to the product of the next digit pair, never to the digit before it is multiplied, and getting the order wrong changes the answer.
- Counting decimal places in only one factor — the product carries the total from both, so 2.37 × 4.6 has three decimal places and not two.
- Skipping a zero row without adjusting the shift — dropping the row is fine only if the next row moves two columns instead of one.
- Treating the nines check as proof — it catches most slips but is blind to transposed digits, so a pass means probably right rather than certainly right.
Related Free Tools From Arb Digital
See the opposite operation laid out step by step on the long division calculator, solve proportions with the cross multiplication calculator, break a number into its prime building blocks with the prime factorization calculator, or work with remainders directly on the modulo calculator. The full free online tools hub lists every mathematics tool we publish.
Frequently Asked Questions
Because it comes from the tens digit of the multiplier. Multiplying by 40 is multiplying by 4 and then by 10, and that final factor of ten is what shifts the row one column left and leaves a zero in the units position.
Yes, and the method does not change. Ignore the decimal points, multiply the digits as whole numbers, then count the decimal places in both factors and give the answer that many places in total.
The product is identical either way, but placing the number with fewer digits underneath gives fewer partial product rows, which means less writing and fewer chances to make an addition error.
It compares the digital roots of the two factors against the digital root of the answer. If the product of the first two does not match the third, there is an arithmetic error somewhere in the layout.
No. It misses any error that leaves the digit sum unchanged, and swapping two digits around is exactly such an error. Treat a pass as strong evidence rather than a guarantee.
Multiply the magnitudes in the ordinary grid and decide the sign separately. Two negatives or two positives give a positive answer, and one of each gives a negative answer.
No, it is the same set of single-digit products arranged differently. The lattice separates them into cells and adds along diagonals, which keeps carries tidy but takes longer to draw than columns.
This page explains a mathematical method for study and for checking your own working. It is not a substitute for showing your method, and it is not medical, legal, or financial advice.
