Extreme Value Theorem: Statement, Proof & Examples

What Is The Extreme Value Theorem?

The Extreme Value Theorem is the result that guarantees a continuous function reaches its highest and lowest values on a closed, bounded interval. Stated precisely:

$$\text{If } f \text{ is continuous on } [a, b], \text{ then there exist } c, d \in [a, b] \text{ such that } f(d) \le f(x) \le f(c) \text{ for all } x \in [a, b].$$

The value $f(c)$ is the absolute maximum and $f(d)$ is the absolute minimum of $f$ on $[a, b]$. Two words in the hypothesis carry the whole theorem, and neither can be dropped:

• Continuous on $[a, b]$: the graph has no jumps, holes, or vertical blow-ups anywhere on the interval.

• Closed and bounded interval $[a, b]$: both endpoints are included, and the interval does not run off to infinity.

The theorem says the two extreme values are attained, meaning some actual input $c$ and some actual input $d$ produce them. That is stronger than saying $f$ is bounded. A function can stay below a ceiling it never touches, and the Extreme Value Theorem rules that possibility out on a closed, bounded interval. For where this sits in the wider subject, see the calculus overview and the related continuity of a function.

Why Do Both Hypotheses Matter For The Extreme Value Theorem?

Drop either hypothesis and the guarantee collapses. The cleanest way to see this is to break one condition at a time and watch an extreme value escape.

Break continuity (a jump). Define $f$ on the closed interval $[0, 1]$ by $f(x) = x$ for $0 \le x < 1$ and $f(1) = 0$. As $x$ climbs toward $1$, the values climb toward $1$ but never reach it, and at $x = 1$ the graph jumps down to $0$. The supremum is $1$, yet no input gives $f(x) = 1$. The interval is closed, but the jump destroys the maximum.

Break closedness (an open interval). Take $f(x) = x$ on the open interval $(0, 1)$. The values get arbitrarily close to $1$ and to $0$, but $x = 0$ and $x = 1$ are not in the domain, so neither bound is attained. Here $f$ is perfectly continuous, yet it attains neither an absolute maximum nor an absolute minimum.

Break boundedness (an infinite interval). Take $f(x) = x$ on $[0, \infty)$. The function is continuous and the left end is included, but the interval runs to infinity, so $f$ has no maximum at all.

Table: What fails when a hypothesis of the Extreme Value Theorem is dropped.

Only the first row satisfies both hypotheses, and only the first row is guaranteed both extremes. Every other row breaks exactly one condition and loses at least one extreme value.

How Do You Prove The Extreme Value Theorem?

The proof splits into two claims: first that $f$ is bounded, then that the bounds are actually reached. Both rest on the same fact that a closed, bounded interval is compact, so any sequence inside it has a subsequence converging to a point still inside it.

Step 1: $f$ is bounded on $[a, b]$. Suppose it is not, so for each whole number $n$ there is a point $x_n$ in $[a, b]$ with $\lvert f(x_n) \rvert > n$. Because $[a, b]$ is closed and bounded, the sequence $x_n$ has a subsequence converging to some point $p$ in $[a, b]$. Continuity forces $f(x_n) \to f(p)$ along that subsequence, a finite number, which contradicts $\lvert f(x_n) \rvert$ growing past every bound. So $f$ must be bounded.

Step 2: the bounds are attained. Let $M$ be the least upper bound of the values of $f$ on $[a, b]$. By the definition of least upper bound, there is a sequence of points $x_n$ in $[a, b]$ with

$$\lim_{n \to \infty} f(x_n) = M.$$

Again the closed, bounded interval hands us a subsequence $x_{n_k} \to c$ with $c$ in $[a, b]$. Continuity gives $f(c) = \lim_{k \to \infty} f(x_{n_k}) = M$, so the maximum value $M$ is reached at the actual point $c$. Running the same argument on $-f$ produces the point $d$ where the minimum is attained. That completes the proof.

The geometric reading is simpler than the symbols. A continuous graph over a closed span cannot approach a ceiling forever without a break in the curve or an escape past the edge of the interval, and a closed, bounded interval forbids both, so the graph is forced to touch its top and its bottom.

What Is The Closed-Interval Method?

The Extreme Value Theorem promises the extremes exist; the closed-interval method is the procedure that locates them. The absolute extremes of a continuous $f$ on $[a, b]$ can occur only at a critical point inside the interval or at an endpoint, so the whole search reduces to a short list of candidates.

1. Confirm $f$ is continuous on $[a, b]$, so the Extreme Value Theorem applies.

2. Find the critical points: solve $f'(x) = 0$ and note any $x$ where $f'(x)$ does not exist, keeping only those inside $(a, b)$.

3. Evaluate $f$ at every critical point and at both endpoints $a$ and $b$.

4. Compare the values. The largest is the absolute maximum; the smallest is the absolute minimum.

The reason interior extremes must be critical points is Fermat's argument: at an interior high or low point the tangent is horizontal, so $f'(x) = 0$ there. Endpoints get their own line on the list because the derivative test never checks them, and the endpoint is very often where the absolute maximum actually lives.

How Do You Apply The Extreme Value Theorem? Worked Examples

Each example checks continuity first, lists the candidates, and compares the values. Every derivative is verified by differentiating the function directly.

Example 1: A polynomial on $[0, 2]$.

Find the absolute maximum and minimum of $f(x) = x^3 - 3x$ on $[0, 2]$.

The function is a polynomial, so it is continuous everywhere, and the Extreme Value Theorem applies on the closed interval. Differentiate and solve for the critical points:

$$f'(x) = 3x^2 - 3 = 0 \quad\Longrightarrow\quad x^2 = 1 \quad\Longrightarrow\quad x = \pm 1.$$

Only $x = 1$ lies inside $(0, 2)$, so the candidates are $x = 0$, $x = 1$, and $x = 2$. Evaluate $f$ at each:

$$f(0) = 0, \qquad f(1) = 1 - 3 = -2, \qquad f(2) = 8 - 6 = 2.$$

Table: Candidate values for $f(x) = x^3 - 3x$ on $[0, 2]$.

Comparing the three values, the largest is $2$ and the smallest is $-2$.

Final answer: absolute maximum $2$ at $x = 2$ (an endpoint), absolute minimum $-2$ at $x = 1$ (an interior critical point).

Example 2: A parabola on $[-1, 3]$.

Find the absolute maximum and minimum of $f(x) = x^2$ on $[-1, 3]$.

The function is continuous, so the theorem applies. Its derivative is $f'(x) = 2x$, which is zero only at $x = 0$, and that point lies inside $(-1, 3)$. The candidates are $x = -1$, $x = 0$, and $x = 3$:

$$f(-1) = 1, \qquad f(0) = 0, \qquad f(3) = 9.$$

The smallest value is $0$ and the largest is $9$.

Final answer: absolute minimum $0$ at $x = 0$ (an interior critical point), absolute maximum $9$ at $x = 3$ (an endpoint).

Example 3: The same function on a wider interval $[-2, 2]$.

Find the absolute maximum and minimum of $f(x) = x^3 - 3x$ on $[-2, 2]$.

The critical points are again $x = \pm 1$, and this time both lie inside $(-2, 2)$, so the candidates are $x = -2$, $x = -1$, $x = 1$, and $x = 2$:

$$f(-2) = -8 + 6 = -2, \quad f(-1) = -1 + 3 = 2, \quad f(1) = -2, \quad f(2) = 2.$$

The absolute maximum $2$ is reached at both $x = -1$ and $x = 2$, and the absolute minimum $-2$ at both $x = -2$ and $x = 1$.

Final answer: absolute maximum $2$, absolute minimum $-2$, each attained at two different points.

Example 3 makes the theorem's real claim visible: it promises that the extreme values exist and are reached, not that they are reached at a single, unique place. Widening the interval from $[0, 2]$ to $[-2, 2]$ kept the same extreme values but changed where and how often they occur.

Why Does The Extreme Value Theorem Work?

The theorem can feel obvious once the picture is right, but its force comes from ruling out two specific escape routes.

• Continuity closes the vertical gaps. A jump or a hole lets a function crowd right up to a value without ever landing on it, the way $f(x) = x$ with a drop at $x = 1$ approaches $1$ but never reaches it. Continuity forbids that near-miss, so a value the graph approaches is a value the graph attains.

• Closedness and boundedness close the horizontal gaps. An open end lets the function creep toward a height that lives just past the missing endpoint, and an infinite interval lets it climb without any ceiling. Including both endpoints and keeping the interval finite blocks both escapes.

• Existence, not location. The theorem is an existence theorem. It certifies that a maximum and a minimum are out there and are reached, which is exactly the licence the closed-interval method needs before it goes looking. Optimisation would be guesswork without that guarantee, because you could never be sure a best value existed to be found.

Read together, the two hypotheses are the two ways a continuous quantity could dodge its own extreme, and the Extreme Value Theorem shuts both. That is why the closed-interval method, and the whole business of maxima and minima, rests on it.

Who Discovered The Extreme Value Theorem?

The theorem looks like common sense, yet proving it rigorously had to wait for a precise definition of continuity and of the real numbers, which arrived only in the nineteenth century.

Two named figures anchor the history:

• Bernard Bolzano (1781–1848, Bohemia) gave the first rigorous proof and, along the way, much of the modern definition of continuity.

• Karl Weierstrass (1815–1897, Germany) reproved it and made it a cornerstone of real analysis, which is why his name is attached to it today.

Where Is The Extreme Value Theorem Used In The Real World?

The theorem is the quiet guarantee behind every problem that asks for a best or worst outcome over a fixed range.

• Engineering and design: finding the maximum stress on a beam or the minimum material for a container over an allowed range of dimensions is an optimization problem, and the Extreme Value Theorem is what promises a best design exists to be found.

• Economics: a firm choosing an output level between zero and full capacity to maximise profit is searching a closed interval, so the theorem assures a profit-maximising level is attained rather than merely approached.

• Physics: the highest point of a projectile or the extreme temperature of a body over a fixed time window is an absolute extreme on a closed interval, guaranteed to be reached.

• Machine learning: training minimises a loss function, and on a closed, bounded region of parameters a minimum is guaranteed to exist, which underpins why the search is well posed.

• Operations and logistics: minimising cost or time over a bounded set of feasible choices relies on the same certainty that a best option is actually attained.

One theorem stands behind them all: whenever a smooth quantity is measured over a closed, finite range, its best and worst values are genuinely reached, so it is worth searching for them.

What Are The Most Common Mistakes With The Extreme Value Theorem?

These four errors account for most lost marks, and each matches a documented student error from AP Calculus reviews, university course notes, and the questions people ask about the theorem.

Applying the theorem when the function is not continuous.

Where it slips in:

A student runs the closed-interval method on a function with a jump or an asymptote inside $[a, b]$, such as $f(x) = 1/x$ on $[-1, 1]$, and reports a maximum.

Don't do this:

Do not assume an extreme value exists before checking continuity. A single break inside the interval voids the guarantee.

The correct way:

Verify $f$ is continuous on the whole closed interval first. If it is not, the Extreme Value Theorem does not apply, and any "extreme" you compute may not exist.

Using the theorem on an open or unbounded interval.

Where it slips in:

A student applies it to $(a, b)$ or to $[a, \infty)$ and expects an absolute maximum, as if $f(x) = x$ on $(0, 1)$ had a top value.

Don't do this:

Do not drop the closed-and-bounded condition. An open end or an infinite interval offers no guarantee at all.

The correct way:

Confirm the interval is closed and bounded, written $[a, b]$ with both endpoints included and finite, before invoking the theorem.

Forgetting to test the endpoints.

Where it slips in:

A student finds the interior critical points, evaluates $f$ there, and stops, never checking $f(a)$ and $f(b)$.

Don't do this:

Do not compare critical points alone. On a closed interval the absolute maximum is often at an endpoint.

The correct way:

Always include both endpoints in the candidate list. In Example 1, the maximum $2$ was at the endpoint $x = 2$, not at the interior critical point.

Confusing the Extreme Value Theorem with the IVT or MVT.

Where it slips in:

A student expects the theorem to hand back the location of the extreme, or mixes it up with the Intermediate Value Theorem or the Mean Value Theorem.

Don't do this:

Do not treat it as a formula that locates or computes the extreme. It only certifies existence.

The correct way:

Keep the roles straight: the Extreme Value Theorem guarantees a maximum and minimum are attained; the closed-interval method then finds them, and the IVT and MVT answer different questions entirely.

Practice Problems On The Extreme Value Theorem

Work each one with the closed-interval method, then check against the answer. Answers are verified.

1. Find the absolute extremes of $f(x) = x^2 - 4x + 1$ on $[0, 3]$.
   (Answer: $f'(x) = 2x - 4 = 0$ at $x = 2$; $f(0) = 1$, $f(2) = -3$, $f(3) = -2$; max $1$ at $x = 0$, min $-3$ at $x = 2$.)

2. Find the absolute extremes of $f(x) = 2x^3 - 3x^2$ on $[-1, 2]$.
   (Answer: $f'(x) = 6x^2 - 6x = 0$ at $x = 0, 1$; $f(-1) = -5$, $f(0) = 0$, $f(1) = -1$, $f(2) = 4$; max $4$ at $x = 2$, min $-5$ at $x = -1$.)

3. Find the absolute extremes of $f(x) = x + \dfrac{4}{x}$ on $[1, 4]$.
   (Answer: $f'(x) = 1 - 4/x^2 = 0$ at $x = 2$; $f(1) = 5$, $f(2) = 4$, $f(4) = 5$; max $5$ at $x = 1$ and $x = 4$, min $4$ at $x = 2$.)

4. Does $f(x) = \dfrac{1}{x}$ attain an absolute maximum on $(0, 1]$?
   (Answer: No. The interval is not closed at $0$ and $f$ is unbounded above there, so the Extreme Value Theorem does not apply.)

5. Find the absolute extremes of $f(x) = \sin x$ on $[0, \pi]$.
   (Answer: $f'(x) = \cos x = 0$ at $x = \pi/2$; $f(0) = 0$, $f(\pi/2) = 1$, $f(\pi) = 0$; max $1$ at $x = \pi/2$, min $0$ at $x = 0$ and $x = \pi$.)

6. On $[-2, 2]$, find the absolute extremes of $f(x) = x^4 - 2x^2$.
   (Answer: $f'(x) = 4x^3 - 4x = 0$ at $x = -1, 0, 1$; $f(-2) = 8$, $f(-1) = -1$, $f(0) = 0$, $f(1) = -1$, $f(2) = 8$; max $8$ at $x = \pm 2$, min $-1$ at $x = \pm 1$.)

Where Should You Go Next After The Extreme Value Theorem?

The theorem is the gateway to optimisation, and several natural doors open from here.

1. Maxima and minima. The full toolkit for classifying and finding extremes, both absolute and local, that the Extreme Value Theorem sets in motion.

2. Critical points. Sharpen the step at the heart of the closed-interval method, where $f'(x) = 0$ or fails to exist.

3. Optimization problems. Apply the guarantee to real word problems, from least material to greatest profit over an allowed range.

If your child is meeting the Extreme Value Theorem for the first time, a live Bhanzu trainer teaches it from the closed-interval picture up, so existence and method feel like one idea, in the Bhanzu math program.