Course contentsShow
← Mathematics
Lesson 1523 of 3,12134. Applications of DerivativesFree lesson

The Second Derivative Test for Extrema

Apply the second derivative test as an alternative method to classify local maxima and minima.

The Second Derivative Test for Extrema

What you'll learn: You'll discover how to use the second derivative as a shortcut to determine whether a critical point is a local maximum, minimum, or neither—without analyzing sign changes around that point.

The Core Idea

You already know the First Derivative Test: find critical points where f'(x) = 0, then check whether f'(x) changes from positive to negative (local max) or negative to positive (local min) around those points.

The Second Derivative Test offers an alternative: once you've found a critical point at x = c (where f'(c) = 0), simply evaluate the second derivative at that point:

  • If f''(c) > 0, the function is concave up at c → you have a local minimum
  • If f''(c) < 0, the function is concave down at c → you have a local maximum
  • If f''(c) = 0, the test is inconclusive—fall back to the First Derivative Test

Why It Works

Think of a valley (minimum) versus a hilltop (maximum). At a valley's bottom, the slope is changing from negative to positive—the graph curves upward (concave up, f'' > 0). At a hilltop, the slope changes from positive to negative—the graph curves downward (concave down, f'' < 0).

The second derivative measures this curvature directly at the critical point, letting you classify the extremum in one calculation instead of testing intervals on both sides.

When to Use It

The Second Derivative Test is most convenient when f''(x) is easy to compute and evaluate. If f''(c) = 0 or undefined, you'll need the First Derivative Test instead. Both methods are valid—choose whichever is simpler for your problem.

Key Takeaway: The Second Derivative Test classifies critical points as local maxima (f'' < 0) or minima (f'' > 0) by checking concavity directly, offering a faster alternative to interval testing when applicable.