The Intermediate Value Theorem turns continuity into an existence guarantee. If a continuous function moves from one output level to another, it must pass through every level between them. The theorem guarantees a solution somewhere, even when an exact formula for that solution is unavailable.
State the hypotheses before the conclusion
If is continuous on the closed interval and N lies strictly between and , then there is at least one in such that equals N.
The interval is part of the claim. You must know continuity throughout it, not merely at its endpoints. The endpoints need not have opposite signs unless the target value N is zero.
The theorem does not say where is exactly, how many such points exist, or whether the function moves monotonically between the endpoints. It supplies existence, not uniqueness or a formula.
Example: prove a root exists
Let This polynomial is continuous on Its endpoint values are and
Zero lies between and 1, so the theorem guarantees some between 0 and 1 with This is a complete existence argument: identify continuity, calculate endpoint values, place the target between them, and state the conclusion.
Saying only “the signs change” omits the continuity requirement. Saying “there is a root at 0.5” goes beyond what the theorem establishes; in fact, not zero.
Continuity and opposite endpoint signs guarantee a zero between 0 and 1, without locating it exactly.
Continuity and opposite endpoint signs guarantee a zero between 0 and 1, without locating it exactly.
Describe this figure
Continuity and opposite endpoint signs guarantee a zero between 0 and 1, without locating it exactly.
Marked points: Negative: filled marker at (0, -1); Positive: filled marker at (1, 1).
Example: the target need not be zero
Suppose a temperature model T is continuous from noon to 4 p.m., with and degrees. The theorem guarantees that the temperature equals 22 degrees at some time between those endpoints.
It does not imply that this occurs at 2 p.m. Midpoint time and midpoint temperature need not correspond unless the model is linear. Nor does it require the temperature to rise throughout the interval; it may fluctuate while still passing through 22.
The same reasoning applies to any target strictly between 18 and 26. Targets outside that range might also occur, but the endpoint information alone does not guarantee them.
Why continuity cannot be skipped
Consider on the interval The endpoint values are and 1, yet the function never equals zero. The function is not continuous on the interval because it is undefined at zero.
A sign change across a vertical asymptote is therefore not proof of a root. The missing continuity allows the graph to move from negative to positive values without passing through zero.
Similarly, a jump function can skip an entire range of output levels. The theorem's power comes precisely from ruling out such breaks.
Narrow a root interval by bisection
For evaluate the midpoint 0.5 and obtain a negative value. Since is positive, continuity guarantees a root in
At 0.75, positive. The root therefore lies in Repeating this process halves a bracket while preserving the existence guarantee.
Bisection locates a root approximately; it does not rely on the graph looking nearly zero. At each step, retain an interval whose endpoints have opposite signs and on which continuity holds.
Existence is different from uniqueness
A continuous function can cross a target level several times. To prove there is exactly one root, you need additional reasoning, such as strict monotonicity on the interval.
For the polynomial above, one can show it is strictly increasing because both and increase. That additional fact gives uniqueness. It is not part of the Intermediate Value Theorem itself.
Try it yourself
Use the theorem to show that has a solution between 1 and 2. Does the theorem alone prove it is the only solution there?
Answer: The continuous function has values 1 and 8 at the endpoints, and 5 lies between them. Thus some in satisfies Uniqueness requires the additional fact that is strictly increasing.
What to remember
Verify continuity on the entire interval, bracket the target with endpoint outputs, and conclude at least one solution. Do not turn an existence statement into an exact location or uniqueness claim.