Why do I need to learn linear approximation if I can just use a calculator?

Linear approximation teaches you how functions behave near a point, which is the conceptual foundation of derivatives and optimization. In real-world applications, it lets engineers and scientists quickly estimate how a system responds to small changes without recomputing from scratch. It is also essential for proving theorems in calculus—the entire theory of derivatives rests on understanding how a function compares to its tangent line.

How do I know if my linear approximation is good enough?

The closer x is to a, the better the approximation. For |x - a| < 0.1 you can usually trust it; beyond that, check the true value if accuracy matters. The curvature of f also matters: if |f''(a)| is large (the function curves sharply), the tangent line diverges faster, so you need x even closer to a.

When f'(a) = 0, the tangent line is horizontal, so L(x) = f(a). This is a valid approximation; it tells you the function is nearly flat at a. If f'(a) is undefined (a corner or vertical tangent), linear approximation does not exist at that point and you need a different approach.

What is the difference between linear approximation and a Taylor series?

Linear approximation is the first-order Taylor polynomial—it uses the value and slope at a. A Taylor series adds higher-order terms (curvature, etc.) to extend accuracy farther from a. Use linear approximation for quick estimates very close to a; use a Taylor series if you need accuracy over a wider range.

Linear Approximation

Q: What if f'(a) = 0?

When f'(a) = 0, the tangent line is horizontal, so L(x) = f(a). This is a valid approximation; it tells you the function is nearly flat at a. If f'(a) is undefined (a corner or vertical tangent), linear approximation does not exist at that point and you need a different approach.

Q: What is the difference between linear approximation and a Taylor series?

Linear approximation is the first-order Taylor polynomial—it uses the value and slope at a. A Taylor series adds higher-order terms (curvature, etc.) to extend accuracy farther from a. Use linear approximation for quick estimates very close to a; use a Taylor series if you need accuracy over a wider range.

Use the tangent line at a known point to estimate a function's value nearby when exact calculations are difficult or unnecessary.

L(x) = f(a) + f'(a)(x - a)

Solve a problem with Linear Approximation

Type the problem. The solver will use Linear Approximation where Linear Approximation is the right tool, and tell you when it is not.

What each symbol means

What Linear Approximation takes

Symbol	Meaning
$L$	The linear approximation function—it is the equation of the tangent line to $f$ at $x = a$ . If you treat $L(x)$ as exact rather than approximate, you ignore the true function's curvature and accumulate error away from $a$ .
$f$	The original function you are approximating. If you use $L$ in place of $f$ for points far from $a$ , the approximation breaks down because the tangent line is flat but $f$ may curve significantly.
$a$	The center point—the fixed $x$ -coordinate where you know both the function value $f(a)$ and the derivative $f'(a)$ . If you confuse $a$ with $x$ or choose $a$ far from your target point, the approximation becomes unreliable.
$x$	The independent variable or input where you want to estimate the function value $L(x)$ . If you plug in a value very far from $a$ , the linear approximation diverges from the true function because you are leaving the neighborhood where the tangent line is accurate.

When to use it

Reach for linear approximation when you need a quick estimate of a function value close to a point where you already know the slope.

Level

Usually taught in: Calculus I

Worked examples

1. Approximate a square root using a nearby integer

Problem

Use linear approximation to estimate

\sqrt{4.1}

using

f(x) = \sqrt{x}

with center point

a = 4

$f(4) = \sqrt{4} = 2$
Evaluate the function at the center point $a = 4$ .
$f'(x) = \frac{1}{2\sqrt{x}}$
Find the derivative of $f(x) = \sqrt{x}$ using the power rule: $f(x) = x^{1/2}$ , so $f'(x) = \frac{1}{2}x^{-1/2}$ .
$f'(4) = \frac{1}{2\sqrt{4}} = \frac{1}{4}$
Evaluate the derivative at the center point.
$L(x) = 2 + \frac{1}{4}(x - 4)$
Apply the linear approximation formula $L(x) = f(a) + f'(a)(x - a)$ .
$L(4.1) = 2 + \frac{1}{4}(0.1) = 2 + 0.025 = 2.025$
Substitute $x = 4.1$ and evaluate.

Answer: $\sqrt{4.1} \approx 2.025$

We chose $a = 4$ because $\sqrt{4} = 2$ is easy to compute and $4.1$ is very close to $4$ . The tangent line at $(4, 2)$ with slope $1/4$ is an excellent approximation for nearby points. The actual value is $\sqrt{4.1} \approx 2.0248...$ , showing that linear approximation is highly accurate when $x$ is very close to $a$ .

2. Approximate an exponential using the tangent line at zero

Problem

Use linear approximation to estimate

e^{0.1}

using

f(x) = e^x

with center point

a = 0

$f(0) = e^0 = 1$
The exponential at zero is always 1.
$f'(x) = e^x$
The derivative of $e^x$ is $e^x$ itself.
$f'(0) = e^0 = 1$
Evaluate the derivative at $a = 0$ .
$L(x) = 1 + 1 \cdot (x - 0) = 1 + x$
Substitute into the linear approximation formula and simplify.
$L(0.1) = 1 + 0.1 = 1.1$
Plug in $x = 0.1$ to get the estimate.

Answer: $e^{0.1} \approx 1.1$

Linear approximation of exponentials near $x = 0$ is remarkably clean because $e^x$ and its derivative are identical, both equaling 1 at zero. This gives the tangent line the simple form $L(x) = 1 + x$ . The true value is $e^{0.1} \approx 1.1052...$ , so the approximation error is less than $0.005$ .

3. Estimate account growth using a known rate of change

Problem

A savings account balance after

t

years is modeled by

A(t) = 1000(1.05)^t

dollars. You know that after 2 years the balance is

\$1102.50

and the account is growing at

\$53.77

per year. Use linear approximation to estimate the balance at

t = 2.1

years.

$L(t) = 1102.50 + 53.77(t - 2)$
Set up the linear approximation $L(t) = A(2) + A'(2)(t - 2)$ using the known balance and growth rate at $t = 2$ .
$L(2.1) = 1102.50 + 53.77(2.1 - 2)$
Substitute $t = 2.1$ into the formula.
$L(2.1) = 1102.50 + 53.77(0.1)$
Simplify the argument: $2.1 - 2 = 0.1$ .
$L(2.1) = 1102.50 + 5.377 = 1107.877$
Compute $53.77 \times 0.1 = 5.377$ and add to the initial balance.

Answer: $A(2.1) \approx \$1107.88$

Linear approximation is invaluable in finance because accountants and investors rarely recalculate the entire formula for near-term projections—they use the current state and known rate of change. This method trades a tiny amount of accuracy for speed and clarity. The exact value from the formula is $1000(1.05)^{2.1} \approx 1107.89$ , so the linear estimate is essentially perfect for a 0.1-year time step.

Common mistakes

Where Linear Approximation usually goes wrong

The mistake	Why it is wrong	The fix
Writing $L(x) = f(a) + f'(x)(x - a)$ or forgetting to evaluate the derivative at $a$ .	The derivative in the formula must be a constant (the slope at one point), not a function of $x$ ; if you use $f'(x)$ , the result is no longer a linear function.	Always substitute $f'(a)$ —that is, evaluate the derivative at the center point—not $f'(x)$ .
Using linear approximation to estimate values far from $a$ without checking whether the result is reasonable.	Linear approximation is only accurate near the center point; as $\|x - a\|$ increases, the tangent line diverges from the actual curve because it ignores the function's curvature.	Restrict linear approximation to neighborhoods close to $a$ (typically $\|x - a\| < 0.1$ or smaller), and for distant estimates, use a better method like Taylor series or compute the exact value.
Leaving $L(x)$ in the form $L(x) = f(a) + f'(a)(x - a)$ and not expanding it to $L(x) = mx + b$ form.	The unexpanded form hides computational errors and makes it harder to spot if you have substituted the derivative correctly; expanding forces you to check your arithmetic.	After plugging in $f(a)$ and $f'(a)$ , distribute and simplify to get the slope-intercept form $L(x) = mx + b$ , then substitute your value of $x$ .

Tips and when to use something else

Linear approximation is just the tangent line equation—if you can write a tangent line in point-slope form, you already know linear approximation.
For small changes in $x$ (roughly $|x - a| < 0.1$ ), linear approximation is nearly exact; error grows fast the farther you move from $a$ .
In applied fields like physics and engineering, linear approximation is preferred for quick estimates because it requires only the current state and rate of change, not the full function formula.
If linear approximation is not accurate enough, use quadratic approximation (the second-order Taylor polynomial) or compute the exact value; do not just hope the error stays small.

Related formulas

Definition of the DerivativeThe Definition of the Derivative finds the instantaneous rate of change at a point using limits; the foundation for all derivative calculations.Power RuleThe Power Rule states that the derivative of x^n is nx^{n-1}, making it the essential tool for differentiating any polynomial expression.Chain RuleThe Chain Rule tells you how to differentiate a composite function—when one function is plugged into another—by multiplying the outer and inner derivatives.Mean Value TheoremThe Mean Value Theorem says there exists at least one point in an interval where a function's instantaneous rate equals its average rate.Critical PointsCritical points mark where a function's derivative is zero or undefined, revealing local extrema and where the graph changes direction.

Frequently asked questions

Why do I need to learn linear approximation if I can just use a calculator?: Linear approximation teaches you how functions behave near a point, which is the conceptual foundation of derivatives and optimization. In real-world applications, it lets engineers and scientists quickly estimate how a system responds to small changes without recomputing from scratch. It is also essential for proving theorems in calculus—the entire theory of derivatives rests on understanding how a function compares to its tangent line.
How do I know if my linear approximation is good enough?: The closer $x$ is to $a$ , the better the approximation. For $|x - a| < 0.1$ you can usually trust it; beyond that, check the true value if accuracy matters. The curvature of $f$ also matters: if $|f''(a)|$ is large (the function curves sharply), the tangent line diverges faster, so you need $x$ even closer to $a$ .
What if $f'(a) = 0$ ?: When $f'(a) = 0$ , the tangent line is horizontal, so $L(x) = f(a)$ . This is a valid approximation; it tells you the function is nearly flat at $a$ . If $f'(a)$ is undefined (a corner or vertical tangent), linear approximation does not exist at that point and you need a different approach.
What is the difference between linear approximation and a Taylor series?: Linear approximation is the first-order Taylor polynomial—it uses the value and slope at $a$ . A Taylor series adds higher-order terms (curvature, etc.) to extend accuracy farther from $a$ . Use linear approximation for quick estimates very close to $a$ ; use a Taylor series if you need accuracy over a wider range.

Need a different method?

The full solver is not scoped to one formula — type any problem and it will pick the method.

Open the math solver

Reviewed 2026-09-18