When we look at the graph of a function, there are sometimes moments where the function reaches a high point or a low point. These are called extreme points. A high point is called a maximum, while a low point is called a minimum. These are examples of extreme points.
At these points, the graph often changes direction. For example, a graph may be increasing before reaching a maximum and then decreasing afterwards. Similarly, it may be decreasing before reaching a minimum and then increasing afterwards. At the very top or bottom of the curve, the tangent line is usually horizontal, which means that its slope is zero.
This connects directly to the first derivative. Since the first derivative gives us the slope of the tangent, an extreme point can often be found by looking for where the first derivative is equal to zero. In other words, we solve to find possible maximum and minimum points, and then investigate what happens around these points to decide whether they are actually maxima or minima.
The important idea is that extreme points are places where a function reaches a local high or low value, and the first derivative helps us find them.
Optional Background – Drawing a Sign Diagram
This is generally less relevant in 3 periods and although it will become relevant below, there are other ways to circumvent having to learn this. However, personally, understanding this gives a seamless way to understand an important point in the theory below (finding extreme points).
Look at the graph of the function
.

Drawing graphs can be tedious. A sign diagram attempts to eliminate the tedious part by focusing only on three things: where the function is positive, where it is negative, and where it is zero. Therefore, we care to first find where the function is 0. We locate those points on a number line, and check in each section if you function is positive or negative by replacing an appropriate value for
.

We know that the line is negative before
and positive afterwards from looking at the graph. Without the graph, we can just replace a value of
that is less than
for the left section and a value that is more than
for the right section.
For example: if
we have
which is negative. On the other hand, if
, we have
which is positive. Exactly as expected. That is, it is much easier and faster to draw a sign diagram that it is a graph.
Example: Determine the signs of the quadratic function
and construct its corresponding sign diagram.
If we wanted to look at the graph of this function, we know it is:

The Role of the Derivative
We have seen already the link between the tangent line and the first derivative. We can use the derivative to understand facts about our function using the following criteria:
Theorem (Local Monotonicity and Tangent Slopes): The connection between a function’s directional behaviour and its derivative at a given point is established through the geometric profile of its tangent line:
- Positive Derivative (
): The slope of the tangent line is positive, meaning the tangent line tilts upwards from left to right. Consequently, the function is locally increasing at that point. - Negative Derivative (
): The slope of the tangent line is negative, meaning the tangent line tilts downwards from left to right. Consequently, the function is locally decreasing at that point.
Let us see how we use this theory.
Example: Determine the signs of the quadratic function
and construct its corresponding sign diagram.
The next example is just to show how when the derivative is a quadratic function (which is harder than something linear), the main idea/method does not change.
Example: Determine where the cubic function
is locally increasing or decreasing by studying the sign of the derivative.
As you can see, there are a few ways to write our conclusions. In the example above we said
, which can also be written as
, or described with words by saying “all numbers between -1 and 1, not including these”.
Finding Extreme Points
The claim above concerns the sign of the derivative. That is, if
or
we get an appropriate conclusion. What about when
? That is the next point we are studying.
Definition (Stationary Point): Let
be a differentiable function and
a point in its domain.
is called a stationary point if:
![]()
: An interior point within the domain of the function where the first derivative vanishes.- Geometric Meaning: The slope of the curve becomes flat, representing a temporary or permanent pause in the function’s vertical growth or decay.
- Classifications: Stationary points can be further categorized as local maxima, local minima, or horizontal inflection points.
From this concept of stationary points, we get the following definition:
Definition (Local Extreme Points): Let
be a real-valued function and
a point in an open interval within its domain.
is called a local extreme point if it satisfies one of the following conditions:
![]()
- Local Maximum: A point where
for all
in a small surrounding interval, geometrically forming a local peak on the curve. - Local Minimum: A point where
for all
in a small surrounding interval, geometrically forming a local trough on the curve. - Necessary Condition: If
is differentiable at an interior extreme point
, then it must also be a stationary point where
.
Finally, we finish the claim we began this page with by giving the real object of our study:
Theorem (The First Derivative Test for Local Extrema): At a stationary point where
, the tangent line is horizontal. The geometric behaviour of the tangent lines around this point determines whether it is a local maximum or minimum:
- From Positive to Negative: If the derivative changes from positive to negative across
, the tangent lines transition from tilting upwards to tilting downwards. The function goes from increasing to decreasing, creating a peak or a local maximum. - From Negative to Positive: If the derivative changes from negative to positive across
, the tangent lines transition from tilting downwards to tilting upwards. The function goes from decreasing to increasing, creating a valley or a local minimum.
In simpler terms, if
before
, then the function was increasing before
. When it changes to
after the point
, then it is decreasing. That is, the function increased until the point
and then decreased, so
is by definition a maximum. The argument for the minimum works analogously.

These special points are referred to as local extrema. We talk about them locally because, on the whole, a graph could have many local maxima or minima but globally they cannot all be maxima/minima unless they are equal. An example of this is the cosine and sine functions, they are periodic and their maximum value is 1 and minimum value is -1.
To find the extreme points of a function
, we follow these steps:
- Find
. - Solve the equation
. - Draw a sign diagram of
and locate all the solutions from the previous step and points of discontinuity of
on it increasingly. - Study the sign of the function
before, after, and in between every one of those points. This will allow you to conclude the behaviour of
. - Conclude which points (if any) are extreme.
Example: Find and classify all local extreme points of the quadratic function
.
Let us now see an example where the derivative ends up being a quadratic function, i.e.: where our original function is cubic. This is just to give a new example where we cannot cheat out the answer by using the vertex.
Example: Find and classify all local extreme points of the cubic function
.
Second Derivative Test
This is a test that requires us to first find the second derivative, so we begin by defining it:
Definition (Second Derivative): The second derivative of a function
is simply the derivative of its first derivative
. It represents the rate of change of the slope of the original function and is denoted as
:
![]()
We call it the second derivative because we are differentiating the original function
twice.
Example: Find the second derivative
of the polynomial function defined by
.
In order to conclude that a point
is a local maximum, we need three conditions:
, i.e.: the tangent line at that point is parallel to the
-axis.
for
, i.e.: the function increases until that point.
for
, i.e.: the function decreases after that point.
Similarly, for
to be a local minimum, we require:
, i.e.: the tangent line at that point is parallel to the
-axis.
for
, i.e.: the function decreases until that point.
for
, i.e.: the function increases after that point.
We have done this by drawing a sign diagram of
, thus making the second derivative irrelevant in this process. However, instead of doing a sign diagram for
we can use the second derivative to draw the same conclusions, as the next claim will show.
Claim (The Second Derivative Test): Let
be a function such that its first derivative equals zero at a critical point
(
), and its second derivative
is continuous in an open interval around
. The classification of this stationary point is determined by the sign of the second derivative at that location:
- If
, then
has a local minimum at
. - If
, then
has a local maximum at
. - If
, the test is inconclusive (the point could be a local maximum, local minimum, or a stationary inflection point).
To use this in an example, we have:
Example: Find and classify all local extreme points of the cubic function
by using the second derivative test.
Exercises
For each of the following four functions, find the coordinates of all stationary points. Differentiate the function to find
, solve the equation
to identify the stationary
-values, and evaluate
at those points to get the full coordinate pairs.
For each of the following real-valued functions, calculate the first derivative, find all stationary points, and classify them as local maxima, local minima, or non-extremes using a sign diagram.
Calculate the second derivative with respect to
for each of the following functions. First evaluate
using standard rules, and then differentiate a second time to determine
. Simplify your answers where possible.
Find the coordinates of all stationary points for each of the following 3 functions and classify them as local maxima or local minima by applying the second derivative test instead of a sign diagram:
- Quadratic Function:
![Rendered by QuickLaTeX.com \[f(x) = x^2 - 8x + 19\]](https://mathematics.lu/wp-content/ql-cache/quicklatex.com-8304d3600c69b49498614f1772fec9e5_l3.png)
- Cubic Function A:
![Rendered by QuickLaTeX.com \[g(x) = x^3 - 3x^2 - 24x + 30\]](https://mathematics.lu/wp-content/ql-cache/quicklatex.com-c67d2a9589e840a5d6bb36826ec22f29_l3.png)
- Cubic Function B:
![Rendered by QuickLaTeX.com \[h(x) = 2x^3 + 9x^2 - 24x - 15\]](https://mathematics.lu/wp-content/ql-cache/quicklatex.com-8c3945a7ad6c5e57c935209bba570cc8_l3.png)


