Percentage Calculator
Find what percent one number is of another, calculate a percentage of a value, or the percentage change between two numbers.
Formula last reviewed 15 September 2026 · How we verify our calculators
45 is this % of 200
Updates live as you type
Frequently asked questions
Divide the percentage by 100 and multiply by the total. 45% of 200 is 0.45 x 200 = 90. The reverse question — what percentage 45 is of 200 — uses a different formula entirely: (45 / 200) x 100 = 22.5%. These two questions look almost identical in English and give completely different answers, which is the most common source of confusion with percentages.
Change % = ((new - old) / old) x 100, and the denominator is always the old value. Going from 45 to 200 is ((200 - 45) / 45) x 100 = +344.4%. Dividing by the new value instead is the classic error and it understates every increase — a salary move from 45,000 to 52,000 is a 15.6% rise against the old figure, but only 13.5% if you wrongly divide by the new one.
Because the base changes. 100 falling by 50% gives 50; a 50% gain on 50 returns only 75. To get back to 100 from 50 you need a 100% gain. The general pattern is that recovering from a fall always requires a larger percentage than the fall itself, and the gap widens sharply at the extremes — a 90% fall needs a 900% gain to break even. This asymmetry is why large drawdowns matter disproportionately in investing.
Divide rather than subtract. To remove 18% GST from an inclusive price of 1,180, compute 1,180 / 1.18 = 1,000, so the tax component is 180. Subtracting 18% of 1,180 gives 212.40, which is wrong, because the 18% was originally calculated on the smaller base. The same applies to recovering a pre-discount price: an item costing 1,500 after 25% off was originally 1,500 / 0.75 = 2,000.
No, they multiply. An offer of 50% off followed by an extra 20% off is not 70% off. The second discount applies to the already-reduced price: 2,000 becomes 1,000, then 800 — a total of 60% off. The combined factor is 1 - (0.5 x 0.8) = 0.60. Successive discounts always total less than the sum of their parts, which is why they are advertised the way they are.
They describe the same change in two different ways and are routinely conflated. If an interest rate moves from 4% to 6%, that is a rise of 2 percentage points but a rise of 50 percent, since 2 is half of 4. Both statements are accurate. Reporting tends to pick whichever sounds more dramatic, so when a figure seems surprising it is worth checking which of the two is being quoted.
Only when the bases are the same size. Scoring 90% on a 10-mark test and 60% on a 90-mark test does not average to 75% — you scored 9 out of 10 and 54 out of 90, so 63 out of 100, which is 63%. Averaging percentages from unequal groups gives every group equal weight regardless of size, which is one of the most common statistical errors in ordinary reporting.
Both measure profit but against different bases. Buying at 80 and selling at 100 gives a profit of 20. As a margin — profit divided by selling price — that is 20%. As a markup — profit divided by cost price — it is 25%. They are equal only when profit is zero, and the gap widens as profit rises. Quoting one while meaning the other is a routine cause of underpricing in small businesses.
No, they compound. Three consecutive years of 10% returns give 1.10 cubed, or 33.1%, not 30%. The effect works against you too: a year of +50% followed by a year of -50% leaves you at 75% of where you started, despite the two figures appearing to cancel. When comparing investments over multiple periods, CAGR is the honest measure because it reflects what actually happened to the money.
Four questions that all look like one
"45% of 200" is 90. "45 is what percent of 200" is 22.5%. "From 45 to 200" is an increase of 344.4%. "From 200 to 45" is a decrease of 77.5%. Same two numbers, four different questions, four unrelated answers.
Almost every percentage mistake comes from answering the wrong one of these — or from assuming that because the numbers are the same, the operations must be related. They are not. A percentage is simply a ratio expressed per hundred, and everything depends on what sits in the denominator.
That is the single idea worth carrying away: *percent of what?* Once the base is identified, the arithmetic is trivial. When people get percentages wrong, it is essentially never the multiplication that failed.
The four operations
Finding a percentage of a number:
Result = (percent ÷ 100) × total
45% of 200 = 0.45 × 200 = 90.
Finding what percentage one number is of another:
Percent = (value ÷ total) × 100
45 out of 200 = (45 ÷ 200) × 100 = 22.5%.
Percentage change between two numbers:
Change % = ((new − old) ÷ old) × 100
The denominator is always the old value — this is where most errors occur. From 45 to 200: ((200 − 45) ÷ 45) × 100 = +344.4%. From 200 to 45: ((45 − 200) ÷ 200) × 100 = −77.5%.
Reverse percentage, recovering the original before an increase or discount:
Original = final ÷ (1 ± rate)
An item costing ₹1,500 after 25% off was originally 1,500 ÷ 0.75 = ₹2,000.
Why increases and decreases are not symmetric
This is the most counter-intuitive property of percentages, and the source of a great deal of confusion in reporting and advertising.
Going from 45 to 200 is a rise of 344.4%. Going back from 200 to 45 is a fall of only 77.5%. The same absolute movement of 155 produces wildly different percentages because the base changed.
The practical consequence: a percentage fall is never undone by the same percentage rise. A stock that drops 50% needs a 100% gain to recover — ₹100 falls to ₹50, and getting from ₹50 back to ₹100 is a doubling. A 20% fall needs a 25% rise. A 90% fall needs a 900% rise.
The same asymmetry applies to discounts and taxes. Adding 18% GST to ₹1,000 gives ₹1,180; removing 18% from ₹1,180 does not mean subtracting 18% of ₹1,180 (₹212.40). It means dividing by 1.18, recovering exactly ₹1,000. Taking the percentage off the larger figure overstates the tax every time.
Worked examples
A discount. A ₹2,000 jacket at 25% off. Discount = 0.25 × 2,000 = ₹500. You pay ₹1,500. To check in reverse: 1,500 ÷ 0.75 = 2,000. ✓
Stacked discounts. "50% off, then an extra 20% off" is not 70% off. The second discount applies to the already-reduced price: ₹2,000 → ₹1,000 → ₹800. That is 60% off in total, not 70. The combined rate is 1 − (0.5 × 0.8) = 0.60.
A marks percentage. 438 out of 500 = (438 ÷ 500) × 100 = 87.6%.
A salary comparison. Moving from ₹45,000 to ₹52,000 a month is ((52,000 − 45,000) ÷ 45,000) × 100 = +15.6%. Computing it against the new salary instead would give 13.5% — the common error, and it understates every raise.
Percentage points versus percent. If an interest rate moves from 4% to 6%, that is a rise of 2 percentage points but 50 percent. Both statements are true and they describe the same event. Headlines routinely conflate them, usually choosing whichever sounds larger.
Where this is used
Shopping — comparing discounts, checking that a "buy two get one free" offer (33.3% off) beats a flat 30%.
Exams and grading — converting marks, computing aggregates, understanding weighting across subjects.
Salary and business — raises, margins, commission, growth rates, market share.
Tax and billing — GST added or removed, TDS deducted, service charges.
Finance — returns, interest, portfolio allocation. Note that investment returns over multiple periods compound rather than add, so three years of 10% is 33.1%, not 30%.
Reading the result
Check the mode matches your question. The most common error on any percentage tool is computing "45% of 200" when the question was "45 is what percent of 200" — the first gives 90, the second 22.5%.
For percentage change, confirm the calculator used your starting value as the base. A negative result means a decrease; the magnitude is bounded at −100% for a decrease (you cannot lose more than everything) but unbounded upward for an increase.
For profit margin specifically, note that margin and markup are different and frequently confused. Buying at ₹80 and selling at ₹100 is a 20% margin (profit ÷ selling price) but a 25% markup (profit ÷ cost price). Quoting one while meaning the other is a routine source of pricing errors in small businesses.
Common mistakes
Using the wrong base for change. Always divide by the original value, not the new one.
Adding stacked percentages. Successive discounts multiply: 50% then 20% is 60% off, not 70%.
Taking a percentage off an inclusive amount. To remove 18% GST, divide by 1.18 — do not subtract 18% of the total.
Confusing percentage points with percent. A move from 4% to 6% is 2 points and 50 percent.
Assuming a fall is reversed by an equal rise. A 50% loss needs a 100% gain to break even.
Averaging percentages of different-sized groups. Scoring 90% on a 10-mark test and 60% on a 90-mark test is not a 75% average — it is (9 + 54) ÷ 100 = 63%. Percentages can only be averaged directly when the bases are identical.
Mixing up margin and markup. They differ whenever the profit is non-zero, and the gap widens as profit rises. ## Doing these in your head
Most everyday percentage questions can be answered without a calculator once you know a few anchors, and the habit is worth building because it catches errors that a calculator will happily produce from a mistyped number.
10% is the anchor. Move the decimal point one place left. 10% of ₹2,340 is ₹234. Everything else builds from there: 20% is double it, 5% is half it, 30% is triple it.
15% is 10% plus half of 10%. For a ₹1,240 bill: 10% is ₹124, half of that is ₹62, so 15% is ₹186.
Reverse the operands when it helps. x% of y always equals y% of x, which is occasionally a large simplification. 8% of 25 is awkward; 25% of 8 is obviously 2.
1% is two decimal places left. Useful for building odd figures — 3% of ₹4,500 is 3 × ₹45 = ₹135.
For discounts, compute what you pay, not what you save. 30% off means you pay 70%, and 70% of ₹800 is ₹560 in one step rather than two.
Sanity-check the magnitude before the digits. If a 20% discount on a ₹3,000 item appears to be ₹60, the answer is out by a factor of ten — 20% is a fifth, so it should be around ₹600. Most calculator errors are order-of-magnitude errors from a missed keystroke, and a rough mental estimate catches essentially all of them.