How to Subtract Fractions With Different Denominators: A 4-Step Method With Worked Examples

Subtracting fractions with different denominators takes four steps: find a common denominator, rewrite both fractions, subtract the numerators, then simplify. As long as you never subtract the bottom numbers, the method works the same whether the denominators are 3 and 4, 8 and 12, or 5 and 7. The reliable version of step one is the least common denominator — the smallest number both denominators divide into. It keeps the arithmetic small and the answer already close to simplest form, which matters most when you subtract mixed numbers or chain several subtractions together.

Why Just Subtracting the Top Numbers Fails

A fraction only makes sense relative to its denominator. Three-quarters means three pieces that are each a quarter of something; two-thirds means two pieces that are each a third of it. Quarters and thirds describe different-size pieces, so counting them together is like adding hours to minutes without converting first. That's the entire reason the denominators must match before any subtraction happens.

The mechanical failure shows up immediately in a wrong answer that feels plausible. Take 3/4 − 2/3. Subtracting across gives 1/1, which equals 1, yet three-quarters of a pizza minus two-thirds of the same pizza is clearly not a whole pizza left over. The correct value is 1/12. The slide from 1 to 1/12 is a useful gut check: two fractions that are both close to 1 should almost cancel, leaving a tiny result.

A quick estimate protects you from this class of error on any test. Both fractions in 3/4 − 2/3 sit between 0.66 and 0.75, so their difference has to be small, under roughly 0.1. If your computed answer is 1, the estimate flags it before you lose points or ship a wrong number in a spreadsheet. Estimation costs about five seconds and catches most denominator errors.

Mixed numbers add a second trap. In 5 1/4 − 2 2/3, the fractional parts (1/4 and 2/3) are separate subtraction problems from the whole numbers, and the fractional subtraction will need borrowing. Handle that department first or convert everything to improper fractions, which removes borrowing from the picture entirely.

Find the Least Common Denominator in Under 30 Seconds

The least common denominator (LCD) is the smallest number that both denominators divide into evenly. For 4 and 3 it's 12, because 12 ÷ 4 = 3 and 12 ÷ 3 = 4 and nothing smaller works. The LCD keeps rewritten numerators as small as possible; any common multiple produces a correct answer, but multiples above the least one force extra simplifying at the end.

There are two fast routes to it. When one denominator divides the other, the larger is the LCD: for thirds and twelfths it's 12, and only the thirds fraction needs converting. When they share a factor, divide it out of the product. For 8 and 12, multiply to get 96, then divide by their greatest common factor of 4 to land on 24. This is the single most valuable shortcut in the whole procedure, because multiplying 8 × 12 = 96 when you don't have to generates bigger numbers, bigger borrowing, and more simplifying.

Here's how the two routes compare on real pairs:

The pattern in that table is worth internalizing: multiplying the denominators always gives a common denominator, sometimes the least one, often not. The GCF shortcut pays off in the 8-and-12 and 6-and-9 rows. For coprime pairs like 5 and 7, multiplying is the least denominator, so no shortcut is needed.

If you're working in a spreadsheet rather than on paper, Excel's LCM function returns the least common multiple directly and GCD handles the divisor step; both are documented in Microsoft's support pages under those exact function names. That's the version-check-free path for anyone doing this at scale — confirm the function name on Microsoft's site if your version is old, since very old Excel releases predate LCM.

Step-by-Step: Subtracting Fractions With Unlike Denominators

Work this sequence in order every time. It handles plain fractions, mixed numbers, and subtraction across a whole number, and it never requires subtracting denominators.

1: Write the problem clearly. For mixed numbers, note the whole number and fractional part separately, e.g., 5 1/4 − 2 2/3. 2: Find the LCD of the two denominators using the division or GCF method above. 3: Rewrite each fraction as an equivalent fraction over the LCD by multiplying top and bottom by the same factor. For 1/4 over 12, multiply by 3 to get 3/12. 4: Borrow if needed. If the first numerator is smaller than the second, take 1 from the whole number and convert it into LCD-sized pieces (1 = 12/12), adding them to the first numerator. 5: Subtract the numerators only. Keep the LCD underneath. For 3/12 − 8/12 after borrowing: 15/12 − 8/12 = 7/12, with the whole numbers already reduced. 6: Simplify by dividing numerator and denominator by their GCF, and convert improper fractions back to mixed numbers if you prefer.

The failure point in that list is step 4, not step 5, in my experience watching students work. The borrowing conversion is where a whole number gets forgotten: 5 1/4 becomes 4 15/12, and the 4 must then be used in the whole-number subtraction (4 − 2 = 2), giving 2 7/12. Dropping back to 5 is the classic slip, and it inflates the answer by exactly 1.

Verify any result by decimal conversion. 5 1/4 = 5.25 and 2 2/3 ≈ 2.6667, so the difference should be about 2.5833. And 2 7/12 ≈ 2.5833. If your fraction answer and your decimal estimate disagree by more than rounding noise, the error is almost always in the borrowing or the numerator subtraction — never in the denominator, which was never touched.

Worked Example 1: 3/4 − 2/3, Step by Step

This is the pair almost everyone meets first, and it's worth doing slowly once. The denominators 4 and 3 share no factor, so the LCD is 4 × 3 = 12. No GCF shortcut applies.

Rewrite both fractions over 12. Multiply 3/4 by 3/3 to get 9/12. Multiply 2/3 by 4/4 to get 8/12. The multiplication factors are always LCD ÷ original denominator: 12 ÷ 4 = 3 and 12 ÷ 3 = 4. That single rule replaces any guesswork about what to multiply by.

Subtract numerators: 9 − 8 = 1, over 12, so the answer is 1/12. Check by decimal: 0.75 − 0.6667 = 0.0833, and 1/12 = 0.0833. The estimate from the opening section — a difference under 0.1 — holds.

Now look at what the wrong method would have produced. Subtracting tops and bottoms gives 1/1 = 1, off by a factor of 12. The ratio of the wrong answer to the right one here is exactly 12, which is the LCD. That's not a coincidence: subtracting denominators instead of finding a common one scales the error by the common denominator itself, which is why the mistake gets dramatically worse as denominators grow.

Worked Example 2: 7/8 − 5/12 With the GCF Shortcut

Here the shortcut earns its keep. The denominators 8 and 12 share the factor 4. Multiply them (96), divide by 4, and the LCD is 24. Rewriting: 7/8 becomes 21/24 (multiply by 3), and 5/12 becomes 10/24 (multiply by 2).

Subtract: 21 − 10 = 11, so the answer is 11/24. The numerator 11 is prime and shares no factor with 24, so this fraction is already simplest form. Decimal check: 0.875 − 0.4167 = 0.4583, and 11/24 = 0.4583.

Had you multiplied straight through to 96, you'd get 84/96 − 40/96 = 44/96, which simplifies to 11/24 after dividing both by 4. Same destination, two extra steps and a larger product to manage. On a timed test with three such problems, that overhead compounds.

The 11/24 result is also the kind of answer that looks "unfinished" to students expecting a tidy fraction, and there's a small practical consequence. In carpentry, 11/24 of an inch lands between the 7/16 (10.5/24) and 1/2 (12/24) marks on a standard tape, roughly 0.458 inch — close enough to 15/32 that a fraction-to-decimal chart is faster than converting by hand on site. Fraction-decimal equivalence charts for this kind of work are published by tool manufacturers and standards bodies; confirm the exact figure on the maker's chart before cutting.

Mixed Numbers: Borrowing When the Top Fraction Is Smaller

Mixed-number subtraction is where most errors live, so treat the whole number as a bank account you withdraw from once. In 7 1/3 − 4 3/4, the LCD of 3 and 4 is 12. Rewrite: 1/3 becomes 4/12, and 3/4 becomes 9/12. You cannot do 4 − 9, so borrow.

Take 1 from the 7, leaving 6, and convert that 1 into 12/12. Add it to 4/12 to get 16/12. Now subtract: whole numbers 6 − 4 = 2, and fractions 16/12 − 9/12 = 7/12. The answer is 2 7/12. Decimal check: 7.3333 − 4.75 = 2.5833, matching 2 7/12.

The alternative most textbooks offer is converting both mixed numbers to improper fractions first. 7 1/3 becomes 22/3, 4 3/4 becomes 19/4, LCD 12, so 88/12 − 57/12 = 31/12, which converts back to 2 7/12. Same answer, no borrowing logic to get wrong, but larger numbers to multiply.

Choose based on the numbers, not habit. If the whole numbers are small and the fractional parts subtract cleanly, borrow directly. If the whole numbers are large (say 43 5/6 − 27 7/8) or you'll be doing many problems, the improper-fraction route keeps everything in one consistent procedure and reduces the chance of forgetting the already-reduced whole number. Both are correct; the borrowing method is faster when it's clean, the improper method is safer when it isn't.

Three Mistakes That Account for Most Wrong Answers

Most subtraction errors cluster into three patterns, and each has a cheap guard. Knowing which one you're prone to shortens practice time considerably.

The first is subtracting denominators. It produces answers that are sometimes whole numbers and always too large. Guard: say the rule out loud before computing — "same bottom, then tops." The guard costs nothing.

The second is forgetting to borrow in a mixed number. You'll see this as a final answer that's exactly 1 too big. Guard: if the second fractional part is bigger than the first (9/12 vs. 4/12), a borrow is mandatory, full stop.

The third is simplifying in the wrong direction or not at all. An answer like 44/96 is correct but unreduced; a student who divides numerator and denominator by different numbers will "simplify" into a wrong fraction. Guard: divide both by the same GCF and check that the new denominator is a factor of the old one. For 44/96, dividing both by 4 gives 11/24 — and 24 is a factor of 96, so the move was legal.

A combined diagnostic covers all three at once: convert your final fraction to a decimal and compare with a rough decimal estimate of the original problem. If the two agree, every step survived the check. If they don't, the error is nearly always in the borrowing or the numerator, and the denominator step can be trusted.

Subtracting Across a Whole Number and Other Edge Cases

Whole numbers lack a visible denominator, which makes them easy to mishandle. In 3 − 1/5, the honest version is 3 = 2 5/5, then subtract: 2 5/5 − 1/5 = 2 4/5. The common error is subtracting 1 from 3 and leaving the fifth untouched, giving 2 1/5, which is exactly 1/5 too large. Borrowing fixes it.

A whole number minus a mixed number is the same drill one level up. 8 − 2 3/7 = 7 7/7 − 2 3/7 = 5 4/7. The 8 must lose 1 to the fractional withdrawal, becoming 7, with 7/7 standing in for the withdrawn whole. If you skip the borrow, you'll try to subtract 3/7 from nothing, which is where the temptation to "just ignore the fraction" creeps in.

Subtracting a fraction from one of its own commons is the tidiest edge case. 1 − 2/5 = 5/5 − 2/5 = 3/5, and the same structure rescues any "1 minus something" problem in probability work: if an event has probability 2/5, the probability it doesn't happen is 3/5, and that complement rule is nothing but this subtraction.

Zero and negative results are legal and worth expecting. 2/5 − 3/5 = −1/5, and 1/3 − 1/2 = 2/6 − 3/6 = −1/6. Don't stop when the numerator would go negative — carry the sign through. In temperature and elevation contexts these answers are meaningful, and a method that halts on negatives is incomplete.

Why Estimation and Improper Fractions Make the Method Faster

Speed in fraction arithmetic comes from estimation, not from bigger computation. Before computing 9/10 − 2/3, note both are close to 1, so the answer should be under 0.3. The true value is 7/30 ≈ 0.2333, which lands inside the estimate. Estimates are cheap insurance and they train number sense in a way that drill alone doesn't.

Improper fractions turn mixed-number subtraction into the same four-step procedure every time. Convert at the start, stay in improper fractions through the LCD and subtraction, convert back at the end. There's no special branch for borrowing, and no whole-number bank account to track. For a student doing twenty homework problems, using one uniform procedure cuts errors more than using the "quicker" method inconsistently.

The two tools combine well in word problems. A recipe calls for 3/4 cup of milk and you have 1/2 cup: subtract to find the shortage, 3/4 − 2/4 = 1/4 cup. Estimate first — 0.75 − 0.5 = 0.25 — and the fractional answer already matches. On a timed test, spending five seconds estimating saves far more than five seconds of rechecking.

If there's a single upgrade to make to your fraction habit, it's this pairing: estimate before, verify by decimal after. The computation in between is the same four steps every time, and the guards catch the borrowing and denominator mistakes that otherwise cost points. None of it requires technology beyond scratch paper, though a calculator's fraction mode can serve as the decimal verifier in a pinch; check the model's manual for how it displays fractional results.

FAQ

Can you subtract fractions with different denominators without finding the LCM?

Yes — multiplying the denominators gives a usable common denominator, and you'll reach the same simplified answer. For 3/4 − 1/6, multiplying gives 24 as the common denominator: 18/24 − 4/24 = 14/24 = 7/12. The LCM of 12 gets you there in one fewer step, cutting the numerators from 18 and 4 down to 9 and 2. Both routes are correct.

What is 1/2 minus 1/3?

1/6. The LCD is 6, so 1/2 = 3/6 and 1/3 = 2/6, and 3 − 2 = 1. Decimal check: 0.5 − 0.3333 = 0.1667. This is the shortest worked example of the full four-step method and a good one to keep in your head as a template.

Why does my mixed-number answer end up exactly 1 too big?

You skipped the borrow. When the first fractional part is smaller than the second (4/12 against 9/12, for instance), the whole number must drop by 1 and that 1 becomes 12/12 added to the fraction. In 7 1/3 − 4 3/4, the correct answer is 2 7/12, not 3 7/12. If your answer is consistently 1 high, this is the cause every time.

Does the least common denominator always work, or is any common multiple fine?

Any common multiple gives a correct answer; the LCD just keeps the numbers smallest and often lands closer to simplest form. With 5/6 − 3/8, the LCD is 24 (20/24 − 9/24 = 11/24), while 48 as a common multiple gives 40/48 − 18/48 = 22/48 = 11/24. The LCD saves the final simplification.

How do you check fraction subtraction quickly?

Convert both fractions to decimals, do the subtraction, and compare with your fractional answer converted to a decimal. For 7/8 − 5/12 the check is 0.875 − 0.4167 = 0.4583 against 11/24 = 0.4583. A mismatch of more than about 0.01 points to a borrowing or arithmetic slip, not to the fraction conversion itself.

Can you subtract a fraction from a whole number directly?

Only by first giving the whole number a denominator. Treat 4 as 3 12/12 if you're subtracting something in twelfths, then subtract the fraction. The shortcut of subtracting the fraction's numerator from the whole number ("4 − 1/3 = 3 2/3") works only in the specific pattern of subtracting a unit fraction; it fails for anything else, so don't rely on it.

What if the numerator goes negative after subtracting?

Carry the negative result and simplify as usual. 1/3 − 1/2 = 2/6 − 3/6 = −1/6. Negative results are correct in temperature, elevation, and balance contexts, and a procedure that stops at zero is incomplete. If your course wants absolute differences, apply the sign at the end rather than mid-calculation.