A proportion calculator solves the equation a/b = c/d when any three of the four values are known. A proportion states that two fractions, or two rates, are equal, and cross multiplication turns that statement into a simple equation you can solve. It is the workhorse method behind recipe scaling, unit conversions, map reading and similar-triangle problems.
This proportions calculator handles the ratio and proportion problems set in most maths courses, and the sections further down cover the statistical meaning of the word as well.
Enter three values, leave the unknown blank, and the calculator finds it, shows the completed proportion, and verifies the answer by reducing both sides to the same decimal. That verification step is exactly how you should check proportion answers on homework and exams.
The word proportion also has a second, unrelated meaning in statistics, where it refers to the fraction of a sample with some characteristic. Both meanings are covered below: cross multiplication first, then sample proportions, confidence intervals and z-tests.
Solving Proportions by Cross Multiplication
If two fractions are equal, their cross-products are equal:
a/b = c/d ⟺ a × d = b × c
This works because multiplying both sides of a/b = c/d by b × d cancels each denominator. To solve for any single unknown:
- Unknown a: a = (b × c) ÷ d
- Unknown b: b = (a × d) ÷ c
- Unknown c: c = (a × d) ÷ b
- Unknown d: d = (b × c) ÷ a
Example: solve 3/4 = 9/d. Cross-multiply to get 3 × d = 4 × 9 = 36, so d = 36 ÷ 3 = 12. Check it: 3 ÷ 4 = 0.75 and 9 ÷ 12 = 0.75. Equal decimals confirm the proportion is true.
Setting Up Proportions from Word Problems
Most proportion mistakes happen during setup, not arithmetic. Keep the same units in the same positions on both sides:
- Recipe: "3 cups of flour make 24 cookies, so how much flour for 40 cookies?" Set up flour/cookies = flour/cookies, giving 3/24 = x/40
- Speed: "A car covers 150 miles in 3 hours, so how far in 5 hours?" Miles over hours on both sides, giving 150/3 = x/5
- Map scale: "1 inch represents 50 miles, so how many miles is 3.5 inches?" Inches over miles, giving 1/50 = 3.5/x, so x = 175
The golden rule: if the numerator on the left is flour, the numerator on the right must also be flour. Mixing positions, putting cookies on top on one side and flour on top on the other, produces a wrong equation that still looks plausible.
Worked Example: Scaling a Recipe
A cookie recipe uses 3 cups of flour to make 24 cookies. How much flour do you need for 40 cookies?
Step 1: Set up the proportion with matching units in matching positions: 3/24 = x/40, flour over cookies on both sides. Step 2: Cross-multiply: 3 × 40 = 24 × x, giving 120 = 24x. Step 3: Divide both sides by 24: x = 120 ÷ 24 = 5.
You need 5 cups of flour. Verify by reducing both sides: 3 ÷ 24 = 0.125 cups per cookie, and 5 ÷ 40 = 0.125 cups per cookie. The per-cookie rate is identical, so the proportion and the answer are both correct.
How to Calculate a Proportion in Statistics
In statistics a proportion is the fraction of a group that has some characteristic, and it is written as a decimal between 0 and 1 rather than as a ratio. How to calculate proportion here is division rather than cross multiplication.
The two you will meet:
- Sample proportion, written p̂ and read "p-hat": p̂ = x / n, where x is the number in your sample with the characteristic and n is the sample size
- Population proportion, written p: the same calculation over the entire population rather than a sample
Example: 320 of 400 surveyed customers say they would recommend a product. The sample proportion is p̂ = 320 / 400 = 0.80, or 80%.
You almost never know the population proportion, which is the entire point. You calculate p̂ from a sample and use it to estimate p, and everything below is about quantifying how good that estimate is.
The calculator above solves a/b = c/d rather than running these statistics, but every formula in the next two sections is worked out fully so you can follow the arithmetic on your own numbers.
Confidence Interval for a Proportion
A single sample proportion is a point estimate, and on its own it hides how much sampling error to expect. A proportion confidence interval puts bounds around it.
The formula has two parts, the estimate and the margin of error:
p̂ ± z × √( p̂(1 − p̂) / n )
The square-root term is the standard error, and z is the critical value for your confidence level: 1.645 for 90%, 1.96 for 95%, and 2.576 for 99%.
Worked from the survey above, with p̂ = 0.80 and n = 400:
- Standard error = √(0.80 × 0.20 / 400) = √0.0004 = 0.0200
- Margin of error at 95% = 1.96 × 0.0200 = 0.0392, about 3.9 percentage points
- 95% confidence interval = 0.80 ± 0.0392, so 0.761 to 0.839
Reported plainly: 80% would recommend, with a 95% confidence interval of 76.1% to 83.9%.
Two things worth knowing. The margin of error shrinks with the square root of the sample size, so cutting it in half takes four times the sample. And the interval is widest when p̂ is near 0.50 and narrowest near 0 or 1, which is why surveys about near-unanimous questions need fewer respondents.
This formula assumes a reasonably large sample. The usual check is that both n × p̂ and n × (1 − p̂) are at least 10. With very small samples or proportions near 0 or 1, the interval can run below 0 or above 1, and an exact method is needed instead.
The 1 Proportion and 2 Proportion z Test
Where a confidence interval estimates a proportion, a z-test asks whether an observed proportion differs from a claim, or whether two proportions differ from each other.
The one proportion z test compares your sample against a hypothesised value p₀:
z = (p̂ − p₀) / √( p₀(1 − p₀) / n )
Note that the standard error uses p₀, the hypothesised proportion, not p̂. Example: a manufacturer claims 55% of users prefer their product, and 232 of 400 surveyed do prefer it.
- p̂ = 232 / 400 = 0.58, and p₀ = 0.55
- Standard error = √(0.55 × 0.45 / 400) = 0.024875
- z = (0.58 − 0.55) / 0.024875 = 1.21
- Two-tailed p-value = 0.228
At the usual 0.05 threshold that is not significant, so the data do not contradict the 55% claim.
The two proportion z test compares two independent groups, and is also written as a z test for two population proportions:
z = (p̂₁ − p̂₂) / √( p̄(1 − p̄) × (1/n₁ + 1/n₂) )
Here p̄ is the pooled proportion, (x₁ + x₂) / (n₁ + n₂), used because the null hypothesis says both groups share one true proportion. Example: 120 of 200 convert in group A and 150 of 300 in group B.
- p̂₁ = 0.60, p̂₂ = 0.50
- Pooled p̄ = 270 / 500 = 0.54
- Standard error = √(0.54 × 0.46 × (1/200 + 1/300)) = 0.045497
- z = (0.60 − 0.50) / 0.045497 = 2.20
- Two-tailed p-value = 0.028
That is significant at 0.05, so the difference between the two groups is unlikely to be sampling noise. This is the calculation behind most A/B test results.
The common mistake is using p̂ rather than the pooled proportion in the two-sample standard error. It inflates z and produces false positives.
Ratio and Proportion: What Is the Difference?
The two words get used interchangeably and they are not the same thing:
- A ratio compares two quantities, written 3:4 or 3/4. It is a single value
- A proportion is an equation stating that two ratios are equal, written 3/4 = 9/12. It has a left side and a right side
So you simplify or scale a ratio, and you solve a proportion. Cross multiplication only applies to proportions, because it needs two ratios set equal to each other.
Every proportion contains two ratios, which is why the two topics are taught together and why "ratio and proportion" is a single chapter in most syllabuses. If you want to simplify a ratio, scale it to a new total, or convert between ratio forms, the ratio calculator handles that directly.
Aspect ratio, as in a 16:9 screen, is a ratio rather than a proportion. Finding the height that matches a given width does set up a proportion though: 16/9 = 1920/x gives x = 1080.
The Golden Proportion
The golden proportion, also called the golden ratio or divine proportion, is the specific ratio where a/b equals (a + b)/a. The solver above doubles as a golden proportion calculator, because it is a proportion in the strict sense. Solving that equation gives an irrational constant usually written with the Greek letter phi:
φ = (1 + √5) / 2 = 1.6180339887...
It is a proportion in the strict sense, which is why the solver above works on it. To find the golden counterpart of any length, set up 1/1.618 = x/length, or simply multiply and divide:
- The larger part of a 100-unit line divided in golden proportion is 100 ÷ 1.618 = 61.8 units, leaving 38.2
- A rectangle 400 units wide is a golden rectangle when its height is 400 ÷ 1.618 = 247 units
The constant has a genuinely interesting mathematical property: the ratio of consecutive Fibonacci numbers converges on it, so 13/8 = 1.625 and 34/21 = 1.619.
The claims about the golden proportion appearing throughout art, architecture and the human body are far weaker than they are usually presented. Most cited examples either do not fit the ratio when measured or were chosen after the fact from many possible measurements. The mathematics is real; the aesthetic universality is largely a 19th-century invention.
Frequently Asked Questions
How do you solve a proportion using cross multiplication?
Multiply the numerator of each fraction by the denominator of the other and set the products equal: a/b = c/d becomes a × d = b × c. Then divide to isolate the unknown. For 3/4 = 9/d, that gives 3d = 36, so d = 12.
What is a proportion in math?
A proportion is an equation stating that two ratios or fractions are equal, such as 3/4 = 9/12. Both sides reduce to the same value, 0.75 here. Proportions model any situation where two quantities change at the same rate, like price per item or miles per hour.
How do you calculate a proportion in statistics?
Divide the count with the characteristic by the total: the sample proportion is p̂ = x / n. If 320 of 400 customers would recommend a product, p̂ = 320 / 400 = 0.80, or 80%. This is a different meaning of the word from a/b = c/d proportions in algebra.
How do you calculate a sample proportion?
Divide the number of items in your sample with the characteristic by the sample size: p̂ = x / n. It is always a decimal between 0 and 1. The population proportion p uses the same arithmetic over the whole population, which is usually unknown and estimated from p̂.
How do you find the confidence interval for a proportion?
Use p̂ ± z × √(p̂(1 − p̂)/n), with z = 1.96 for 95% confidence. For p̂ = 0.80 and n = 400, the standard error is 0.02 and the margin of error is 0.0392, giving an interval of 0.761 to 0.839. Check that both n × p̂ and n × (1 − p̂) are at least 10 before relying on it.
What is the formula for a two-proportion z-test?
z = (p̂₁ − p̂₂) ÷ √(p̄(1 − p̄) × (1/n₁ + 1/n₂)), where p̄ is the pooled proportion (x₁ + x₂)/(n₁ + n₂). With 120 of 200 and 150 of 300, p̄ = 0.54 and z = 2.20, giving a two-tailed p-value of 0.028. Using p̂ instead of the pooled proportion is the most common error.
How do I check whether a proportion is true?
Use either test. Cross-multiply and see if the products match: for 3/4 = 9/12, both 3 × 12 and 4 × 9 equal 36. Or divide each fraction and compare decimals, which gives 0.75 on both sides. If either test passes, the proportion is true.
How do I set up a proportion from a word problem?
Identify the two quantities being compared and keep each in the same position on both sides. If 150 miles pairs with 3 hours, write miles over hours on both sides: 150/3 = x/5. Label the units while setting up, since mismatched positions are the most common source of wrong answers.
What is the difference between a ratio and a proportion?
A ratio compares two quantities, such as 3:4, and is a single value. A proportion is an equation saying two ratios are equal, such as 3/4 = 9/12. You simplify or scale a ratio and you solve a proportion, and cross multiplication only applies to proportions.
What is the golden proportion?
The ratio where a/b equals (a + b)/a, which works out to φ = (1 + √5)/2 = 1.618034. Dividing a 100-unit line in golden proportion gives parts of 61.8 and 38.2. The mathematics is real, including the link to Fibonacci numbers, but the claims about it appearing throughout art and the human body are much weaker than usually presented.