What Is a Counterexample in Math? One Case Is Enough

If you have ever wondered what is a counterexample in math, start with this claim: “Every odd positive integer is prime.” The number 9 meets the conditions: it is positive and odd. But 9 = 3 × 3, so it is not prime. That makes 9 a counterexample. You do not need to check any other numbers to decide whether the original claim is true.
What is a counterexample in math and why does one suffice?
A counterexample is a case that meets a universal claim's conditions but fails its conclusion. A universal claim says something is true for every member of a specified set. Words such as all, every, and always often signal this kind of claim. So does a statement such as “If an integer is odd and positive, then it is prime,” when it is meant to apply to any integer.
To check a possible counterexample, ask two things: Does it belong to the group the claim is about? Does it make the conclusion false? The number 4 is not prime, but it does not disprove the claim about odd positive integers because it is even. The number 9 passes both checks. The word every leaves no room for even one exception.
Pay attention to the difference between every and some. The number 9 does not disprove “Some odd positive integers are prime.” In fact, 3 proves that statement because it supplies exactly what some asks for: at least one case that works.
Why a long list of successes is not a proof
Consider the claim “For every positive integer n, the product n(n + 1) is less than 1,000,000.” For n = 1, the product is 2. For n = 999, it is 999,000. If you tested all 999 positive integers from 1 to 999, every test would support the claim. But at n = 1,000, the product is 1,001,000, which is greater than 1,000,000. The first 999 results are correct. The mistake would be treating those results as evidence about numbers you never tested.
Checking examples can prove a universal claim when the allowed set is finite and you actually check every case. If the product claim were restricted to positive integers from 1 to 999, testing all of them would settle it. You could also prove the restricted claim without listing them: in that range, n is at most 999 and n + 1 is at most 1,000, so the product is at most 999,000.
Examples are useful even when they do not prove a claim. They can suggest a pattern or help you spot a mistake in your reasoning. But if some cases remain unchecked, you still need an argument that covers them, not just a longer list of successes.
Start with zero and negative numbers
Test zero if the claim allows it. Zero often makes terms disappear, and it can turn a strict inequality into an equality. Take the claim “For every real number x, x2 + 1 > 1.” When x = 0, the left side equals 1, so the strict inequality fails. If you change > to ≥, you get a true claim: every real square is nonnegative, so adding 1 gives a result of at least 1.
Negative numbers are especially useful when a claim involves multiplication or order. Suppose someone says, “For all real numbers a, b, and c, if a < b, then ac < bc.” Try a = 1, b = 2, and c = -1. The condition 1 < 2 holds, but ac = -1 and bc = -2. Since -1 > -2, the conclusion fails. This example shows why the sign of the multiplier matters: multiplying both sides of an inequality by a negative number reverses its direction.
Keep the allowed set in view. Neither zero nor a negative number can serve as a counterexample to a claim restricted to positive integers. In that setting, try 1 first. For instance, “Every positive integer has at least two distinct positive divisors” fails at 1, whose only positive divisor is itself.
Check endpoints and places where the behavior changes
Boundary cases sit at or near the edges of the allowed set, or near a point where an expression changes behavior. For the claim “If 0 ≤ x ≤ 1, then x2 < x,” the midpoint x = 0.5 works because 0.25 < 0.5. Both endpoints fail: at x = 0 and at x = 1, the square equals the original number. If the claim instead says 0 < x < 1, those endpoints no longer count. The revised claim is true because multiplying x < 1 by the positive number x gives x2 < x.
Not every useful boundary is an endpoint. For “For every real x, (x - 3)2 > 0,” there are no endpoints to check. Instead, look for the point where the expression inside the square becomes zero. At x = 3, the square is zero, so the strict inequality fails. When you see a strict inequality, finding where the two sides could be equal is often faster than trying values at random.
Write a counterexample someone else can check
A short disproof should show both the condition and the failure. For “Every integer greater than 10 is a sum of two even integers,” you could write: “The integer 11 is greater than 10. But the sum of two even integers is always even, and 11 is odd. Therefore 11 cannot be written as such a sum.” Just writing “11” gives the right candidate without showing why it works.
For more practice, use the IYMC training and practice problems and the guide to showing your work in math to practice making each check explicit.
Try this claim on paper: “For every integer n greater than or equal to 2, n2 > 2n.” Start with the smallest allowed value. Write one line showing that it meets the condition and one line comparing the two sides.
Frequently asked questions
Can a counterexample use more than one number?
Yes. One case can be a pair or a group of values if the claim has several variables. All the stated conditions must hold for that case.
Can equality disprove an inequality?
Equality disproves a strict inequality, such as x² > x at x = 1. It does not disprove an inequality that allows equality, such as x² ≥ x.
Does failing to find a counterexample mean the claim is true?
No. Unless you have checked every allowed case, you may simply have missed an exception. To prove the claim, give an argument that covers every allowed case.



