U-substitution is the reverse of the chain rule. It helps when an integral contains a function inside another function and also contains something close to the derivative of the inside function. The goal is to replace the complicated inside expression with a simpler variable, usually called \(u\).
When to try u-substitution
Look for a nested expression such as \((x^2 + 1)^5\), \(\sin(3x)\), \(e^{4x}\), or \(\ln(x)/x\). If the derivative of the inside expression appears elsewhere in the integrand, substitution is likely a good fit.
How to choose u
Choose \(u\) as the inside expression. For example, in the integral of \(2x(x^2 + 1)^5\), choose \(u = x^2 + 1\). Then \(du = 2x\,dx\). The remaining integral becomes \(\int u^5\,du\), which is much easier.
Worked example
Problem: integrate \(2x(x^2 + 1)^5\,dx\). Let \(u = x^2 + 1\). Then \(du = 2x\,dx\). The integral becomes \(\int u^5\,du\). That equals \(u^6/6 + C\). Substitute back to get \((x^2 + 1)^6/6 + C\).
Example with a constant adjustment
For the integral of \(\cos(3x)\,dx\), let \(u = 3x\). Then \(du = 3\,dx\), so \(dx = du/3\). The integral becomes one third of the integral of \(\cos(u)\,du\), which is \(\sin(3x)/3 + C\).
Common mistakes
Students often forget to replace \(dx\), or they choose \(u\) but leave part of the original \(x\) expression behind. Another common issue is forgetting to substitute back into the original variable at the end.
Building confidence with substitutions
Substitution becomes easier when you train yourself to look for pairs. One part of the integrand should be the inside expression, and another part should resemble its derivative. The match does not have to be perfect because constants can be adjusted, but the variable pieces need to line up. If you choose \(u\) and still have an extra \(x\) expression that cannot be rewritten in terms of \(u\), the substitution probably needs to be changed.
Use internal practice routes instead of jumping randomly between topics. Start with the indefinite integrals guide to review direct antiderivative rules, then return here for composite expressions. If the substitution produces a definite integral with bounds, also compare the setup with Definite vs Indefinite Integrals, because bounds can either be converted to \(u\)-values or handled after substituting back.
Common classroom problems are designed so the derivative clue is visible. In more realistic expressions, the clue may be hidden by algebra. Expanding, factoring, or rewriting radicals as powers can reveal the inside function. This is why simplification before choosing a method matters.
How to check your answer
Differentiate your final answer. If you get the original integrand, the substitution worked. You can also compare with the Integral Calculator, but the derivative check is the strongest habit.
Study checklist
For every substitution problem, write three lines clearly: the choice of \(u\), the expression for \(du\), and the transformed integral. If one of those lines is missing, the solution is harder to audit. In definite integrals, also decide whether to change the bounds into \(u\)-bounds or substitute back before evaluating.
When practicing, keep a small list of inside functions you notice often: quadratic expressions, linear angles inside trig functions, denominators inside logarithms, and powers of binomials. Seeing these patterns repeatedly makes substitution feel less like guessing and more like recognizing the chain rule in reverse.
Practice prompts
Create three substitution problems with the same inside expression. For example, use \(x^2+1\) inside a power, inside a square root, and inside a denominator. Then ask whether the outside factor provides the derivative \(2x\). If it does, substitution is likely direct. If it does not, decide whether a constant adjustment is enough or whether the method fails.
Also practice explaining your choice of \(u\) in words. A strong explanation might say, I chose the repeated inside expression because its derivative appears elsewhere in the integrand. This sentence forces you to check the structure instead of choosing \(u\) by habit.
Teacher-style review questions
What expression is being simplified by the substitution? Where does its derivative appear? Is a constant adjustment needed? Can every remaining part of the integrand be written in terms of \(u\) and \(du\)? These questions make substitution transparent. They also explain why some tempting substitutions fail even though they look reasonable at first glance.