AP Calculus integration using substitution, often called u-substitution, is the single most tested antidifferentiation technique on the exam, and it is also the technique that loses candidates the most points when they execute it half-correctly. The College Board awards credit row by row: one row for choosing a workable substitution, one row for converting the differential, one row for producing an antiderivative in the new variable, one row for back-substituting or converting bounds, and one row for the constant of integration or the final numerical value. Every one of those rows can be forfeited by a single sloppy step, and most candidates lose at least two. The work below maps each rubric row to the moves that earn it, isolates the AB versus BC differences, and gives a concrete six-minute template for the FRQ form where substitution shows up most often.
What the rubric actually rewards when you write a u-substitution
Read any AP Calculus FRQ that contains a u-substitution and you will see the same four-to-five-row pattern underneath the surface. The first row is the choice of u. The reader is not asking whether u is the only possible choice; the reader is asking whether your chosen u, paired with du, makes the remaining integral a standard form the syllabus covers. Picking u = 2x + 1 when the integrand is (2x + 1)^7 · 2 dx is the canonical move, and it earns the substitution row cleanly. Picking u = x in the same problem technically leaves the integral unchanged, which is technically correct but scores zero on the substitution row because no simplification has happened.
The second rubric row is the differential conversion: writing du = 2 dx, or dx = du / 2, and rewriting the entire integrand in u alone. Candidates who leave x terms inside the rewritten integral forfeit this row even if the rest of the work is perfect. The third row is the antiderivative in u. Here the grader is looking for a clean primitive such as u^8 / 8 or ln|u|, written with correct constants absorbed. The fourth row is the back-substitution (for indefinite integrals) or the bounds conversion (for definite integrals), and the fifth row is the constant of integration or the final numerical value.
Notice the structural point: the rubric is rewarding algebraic hygiene, not cleverness. A candidate who writes a slightly unusual but still valid substitution — for instance u = 1 + x^2 on ∫ 2x cos(1 + x^2) dx — earns the same substitution row as someone who picked the textbook choice, provided the rest of the chain is correct. What costs the row is mixing variables: writing u^8 / 8 + C with no reference back to x, or evaluating u = 2x + 1 at x = 0 and x = 3 after the integral is already in u form.
The antiderivative row: dx, du, and the constant that quietly lives inside
AP Calculus readers see the same five errors on the antiderivative row in almost every administration. Number one: candidates forget to convert the dx at all, so the antiderivative is written in mixed form. ∫ (2x + 1)^7 · 2 dx becomes (2x + 1)^8 / 8 + C, which is correct in x by accident, but the rubric row explicitly asks for the substitution step. The reader cannot award the substitution row if no substitution was visibly performed. The fix is mechanical: always write du = 2 dx as its own line, then substitute on the next line.
Number two: the constant of integration appears too early, before the back-substitution. u^8 / 8 + C is fine as a mid-solution line, but the +C is a property of the final antiderivative in x. Some readers will accept an early +C; others will mark it as ambiguous. The defensive move is to write C only on the line that returns to x. Number three: the inner derivative is dropped. ∫ cos(x^2) dx looks like a u-substitution problem with u = x^2, but the resulting du = 2x dx means the integral becomes (1/2)∫ cos(u) du. Candidates who forget the factor of 1/2 lose the antiderivative row even though their chain of thought was otherwise correct.
Number four: the back-substitution is performed but the algebra is wrong. (2x + 1)^8 / 16 + C is a frequent error from conflating 1/2 with 1/8 on the way back to x. Number five: the final answer is left in u. u^8 / 8 + C standing alone forfeits the back-substitution row. The defensive template is:
- Line 1: state u and du.
- Line 2: rewrite the integrand in u only, including dx.
- Line 3: write the antiderivative in u.
- Line 4: back-substitute to x.
- Line 5: append +C on the x-line.
Adopting this template costs about twenty seconds per integral, and on the FRQ it is the difference between a 3 and a 5 on the row.
Definite integrals: bounds, no-+C, and the back-substitution that disappears
Once the limits of integration are present, the rubric shifts. For definite integrals, there is no +C row at all; the constant is absorbed by the evaluation. More importantly, the candidate has two legal paths. Path A is to convert the bounds from x to u immediately after the substitution and never return to x. Path B is to back-substitute to x first and then evaluate at the original bounds. The grader accepts either path, but the rubric points are not symmetric: a candidate who mixes the two paths — say, by writing the antiderivative in u, then evaluating at the x-bounds — forfeits both the bounds row and the evaluation row.
The cleanest Path A example: ∫₀^3 2x(1 + x²)^4 dx. Set u = 1 + x²; then du = 2x dx; the bounds become u(0) = 1 and u(3) = 10; the integral becomes ∫₁¹⁰ u^4 du; the antiderivative is u^5 / 5 evaluated from 1 to 10; the answer is (10^5 − 1) / 5 = 99,999 / 5. Every line is in u. There is no back-substitution. The constant of integration is unnecessary because definite integration produces a number, not a family.
The cleanest Path B example: same problem, but after writing u^5 / 5 the candidate writes (1 + x²)^5 / 5 and evaluates at x = 0 and x = 3. This is also acceptable. The dangerous move is to evaluate the u-form antiderivative at the x-bounds, e.g. (1 + 0²)^5 / 5 − (1 + 3²)^5 / 5 but written as u^5 / 5 at the top. That is the row-killer, because the rubric cannot tell which variable the bounds refer to.
On the BC exam, definite u-substitution problems are often embedded inside a larger accumulation or area problem where the candidate must also identify the integrand, set up the limits, and interpret the answer. The substitution row is still there, but it is one of four or five rows on the FRQ, and graders are explicitly told not to award the substitution row twice. In practice, the rule means: if you correctly identified a workable u and your antiderivative is right, the substitution row is yours even if the back-substitution line is a little rough. If your antiderivative is wrong, the substitution row still scores if the choice of u is reasonable. That is the safety net most candidates do not realise they are sitting on.
Where AP Calculus AB stops and BC begins on substitution
On AB, substitution is essentially limited to the form ∫ f(g(x)) g'(x) dx, with a polynomial or simple trigonometric inside. Typical AB stems ask for the antiderivative of (3x² + 1) · cos(x³ + x) or 6x · e^(x²). The skill being measured is recognition: did the candidate see that the inner function and its derivative are both present, and did they pick the right u? The back-substitution is one step. The constant of integration is required. No definite-integral accumulation language is needed.
On BC, the same recognition skill is present but layered with two additional demands. First, BC candidates are expected to handle substitution inside definite integrals without back-substitution, choosing bounds conversion as the cleaner path on volume, area-between-curves, or probability-density problems. Second, BC candidates encounter substitution on logistic and exponential-model FRQs where the substitution is not a polynomial — for instance, integrating 2^x · ln 2 by recognising 2^x as the derivative of 2^x / ln 2, or integrating e^(2x) sin(e^x) · cos(e^x) by setting u = sin(e^x) and handling the inner chain twice. These problems test whether the candidate can stack the chain rule, not just apply it.
The defensive rule for BC candidates: if the integrand contains a transcendental function composed with x, ask whether the derivative of the inside is sitting as a multiplicative factor before reaching for any other technique. Most BC substitution points are awarded on problems where the candidate could have been tempted into integration by parts or a trig identity, but the substitution is the only one that works in under four minutes. The grader's mental model is, "Did the candidate see the structure?" The structure is inside-function derivative × outside-function-of-inside. Once you see that pattern, the rubric rewards the recognition row even if the arithmetic on the next line is slightly off.
Worked example: a six-minute FRQ walkthrough
Consider a typical AB-style FRQ stem: Let f(x) = ∫₀^x (t² + 1)⁻¹ dt. Find f(2) and write the equation of the tangent line to f at x = 2. A six-minute budget that hits every rubric row looks like this.
Minute one: recognise that (t² + 1)⁻¹ is not elementary on its own; do not attempt to antidifferentiate symbolically. Instead, this problem usually has a calculator-active component, but if the FRQ demands an analytical form, the candidate needs to perform a trigonometric substitution, which is BC territory. If we are on AB, the problem is likely to be evaluated numerically using the calculator; if we are on BC, the candidate should try t = tan θ, giving dt = sec²θ dθ and t² + 1 = sec²θ, so the integrand collapses to dθ. The rubric gives credit for setting up that substitution.
Minute two: state the substitution cleanly. Let t = tan θ, dt = sec²θ dθ. The integrand (t² + 1)⁻¹ dt becomes dθ. That line alone earns the substitution row. Minute three: integrate. ∫ dθ = θ + C = arctan(t) + C. This earns the antiderivative row. Minute four: back-substitute. f(x) = arctan(x) + C. Apply the initial condition f(0) = 0 to get C = 0. Minute five: evaluate. f(2) = arctan(2). Minute six: differentiate to get the tangent slope, f'(2) = (4 + 1)⁻¹ = 1/5, and write the tangent line, y − arctan(2) = (1/5)(x − 2).
The point of the walkthrough is not the specific answer; it is the row-by-row mapping. Each of the six minutes corresponds to one rubric row, and the candidate who runs out of time at minute four still has four of the five rows. On the AP exam, partial credit is the floor and the ceiling at the same time: it is the worst-case outcome and the realistic best-case outcome. A six-minute template turns partial credit into a planning tool.
Common pitfalls and how to avoid them
Five pitfalls produce the majority of the lost points on u-substitution FRQs. The first is choosing a u that does not simplify the integrand. Candidates sometimes pick u equal to the outermost function (for instance u = sin(x²) on a problem whose structure is actually u = x² inside the sine). The defensive move: always ask, "If I differentiate my u choice, does the derivative appear elsewhere in the integrand?" If the answer is no, the choice is wrong.
The second pitfall is forgetting the dx → du conversion and writing a final answer in u that secretly contains x. The defensive move: after the substitution, scan the rewritten integral and confirm there is no x left. The third pitfall is over-substituting, picking a u that requires two more substitutions to finish. On the AP exam, the substitution is supposed to take one step. If you find yourself planning a second substitution, you have probably picked the wrong u.
The fourth pitfall is dropping the constant of integration on indefinite integrals, or worse, including it on definite integrals. The defensive move: read the integral. If there are no bounds, the +C row is one of the rubric rows. If there are bounds, the +C row is not there. The fifth pitfall is mixing paths on a definite integral. The defensive move: choose Path A (convert bounds) or Path B (back-substitute) before you start writing, and stay on that path to the end.
Preparation strategy: how to train the substitution row specifically
Candidates who want to lift their substitution score should practise against the rubric directly rather than against a textbook chapter. Take any released FRQ that contains a u-substitution, cover the answer key, and grade your own work using the five-row template. Score one point per row. After ten problems, you will see a pattern: most candidates lose the back-substitution row and the constant row, not the substitution row. Knowing which row you personally lose lets you focus the next week of practice on that row, not on re-reading the whole chapter.
Within the AP Calculus syllabus, the relevant Course and Exam Description units are Differentiation (Unit 3) for the chain-rule recognition, and Integration (Units 6–8) for the antiderivative application. The most efficient preparation sequence is to drill the recognition step on twenty short stems — given an integrand, write down the correct u and du — and then move to the full write-up. The recognition step is what is tested on the multiple-choice section; the full write-up is what is tested on the FRQ. The two skills are different, and the rubric rewards them on different rows.
For time budgeting on the FRQ, the realistic per-problem target is four to six minutes on a substitution-only problem, and eight to ten minutes on a substitution-embedded problem where you also need to set up an accumulation or area context. Candidates reading this should practise to a stopwatch, not to a chapter-end problem set. The exam rewards the candidate who can produce five clean rows in six minutes, not the candidate who produces one perfect answer in twenty.
Comparative scoring: what a 3, a 4, and a 5 look like on a substitution row
The College Board reports scores on a 1–5 scale, but the rubric underneath is a row-by-row tally. On a typical four-point substitution problem, a score of 3 usually means the candidate earned the substitution row, the antiderivative row, and the evaluation row, but forfeited the back-substitution row by leaving the answer in u. A score of 4 means the candidate earned everything except the constant of integration on an indefinite integral. A score of 5 means the candidate produced a clean five-row write-up with no algebraic slips.
| Rubric row | What the grader looks for | Common way the row is lost | Defensive move |
|---|---|---|---|
| Choice of u | u and its derivative both appear in the integrand | u chosen without checking the derivative | Differentiate u, scan for it in the integrand |
| Differential conversion | dx rewritten in terms of du, no x left | dx left in the rewritten integral | Re-scan the rewritten line for any x |
| Antiderivative in u | Correct primitive in the new variable | Dropped a constant factor, e.g. 1/2 | Compute the primitive, then re-differentiate to check |
| Back-substitution or bounds | Return to x, or convert the bounds to u | Mixing u-form antiderivative with x-form bounds | Pick Path A or Path B and stay on it |
| Constant of integration / final value | +C on indefinite; correct number on definite | +C missing on indefinite, or present on definite | Read the integral before writing the final line |
Notice that a 3, a 4, and a 5 differ by only one or two rows. That is the structural feature of the rubric. A candidate who can hold all five rows for sixty percent of the substitution problems they encounter will land at a 4 or a 5 overall, because the other FRQ rows on the paper (units, setup, justification) compound in the candidate's favour. A candidate who can hold only three rows will land at a 3 even if their raw content knowledge is strong, because partial credit on this exam is row-by-row, not topic-by-topic.
Conclusion and next steps
AP Calculus integration using substitution is not a content problem; it is a row-by-row writing problem. The candidate who treats the rubric as five discrete checkpoints, and who practises against the rubric rather than against the textbook, will outscore the candidate who re-reads the chapter one more time. The single most useful drill between now and the exam is to take ten released FRQs containing a u-substitution, score yourself against the five-row template, and identify which row you personally lose most often. Then drill that row, not the whole problem. If the back-substitution row is your weak point, drill fifteen pure back-substitution problems in twenty minutes. If the differential conversion is your weak point, drill the dx → du step in isolation. The exam rewards targeted repair, not general review.
AP Courses' one-to-one AP Calculus AB and BC programmes work through each candidate's specific row-loss pattern on u-substitution FRQs, mapping their free-response work against the rubric and turning a 5 target into a row-by-row preparation plan tied to the actual College Board scoring guide.