On the AP Calculus exam, an integral of a composite function is the moment where most of the syllabus converges at once. A student has to recognise the outer function, identify an inner function whose derivative is (up to a constant) sitting in the integrand, perform the substitution, evaluate the antiderivative, and then translate the result back into the original variable if the answer is required in x rather than in u. Each of those steps maps onto its own rubric row, and a single missing piece — the back-substitution, the differential, the constant of integration — can cost a full point on a free response question even when the rest of the work is sound. This article walks through how the AP Calculus rubric actually scores these problems, which question types appear in the multiple-choice and free-response sections, and what a deliberate preparation strategy looks like in the weeks before the exam.
What counts as an integral of a composite function in the AP Calculus syllabus
In the language of the College Board's course description, an integral of a composite function is any antiderivative problem whose integrand can be written as f(g(x))·g'(x) for some pair of differentiable functions f and g. The standard example is the sine-of-cosine pattern, where the integrand is sin(x)·cos(x) and the antiderivative is −cos²(x)/2 plus a constant. That one example, more than almost any other, is the canary in the coal mine for a student's preparation: it tests whether the student sees structure rather than pattern-matches a formula.
The AP Calculus AB syllabus places this technique inside Unit 6, "Integration and Accumulation of Change," and treats it as a required skill. The AP Calculus BC syllabus keeps the same requirement and folds it into Unit 8, "Applications of Integration," alongside other substitution-heavy applications. In both courses, a student is expected to handle two broad families: integrals whose substitution is transparent because the integrand visibly contains a function and (a constant multiple of) its derivative, and integrals whose substitution must be chosen strategically because the integrand only resembles the chain rule form after a small algebraic manipulation.
The first family is what shows up on the early multiple-choice items. A stem might present the integral of 2x·cos(x²) dx and ask for a particular antiderivative. The second family is what tends to land on the free response section, where the integrand is something like x·√(x²+1), forcing the student to recognise that the derivative of the inner expression is 2x, then divide by 2 to align the differential, and only then perform the substitution. The BC exam adds a third layer, where the substitution can produce a definite integral that the student must evaluate in u-space, sometimes in combination with a different technique such as integration by parts.
For a candidate reading this, the practical implication is that the topic is not optional. It appears in roughly one in every four or five free response questions on the AB exam and in slightly more on BC, and a single missed back-substitution on the 2018-style accumulation problems cost several thousand candidates a full point when the rubric was first released. Treat composite-function integration as a high-yield, high-penalty topic, and plan accordingly.
The u-substitution row on the AP Calculus FRQ: what the rubric actually requires
The free response scoring guidelines for AP Calculus treat a u-substitution problem as a sequence of discrete, scorable rows. The graders are not looking for a beautifully written narrative; they are looking for a sequence of auditable statements, each of which can be checked against the rubric. On a typical 9-point FRQ, a composite-function integral question will claim three or four of those points, distributed across the following rows.
The first row is the identification of u. The student must write the substitution explicitly, and the chosen u should be the inner function — for example, u = x² + 1 — rather than a vague "let u be the inside." A surprisingly common error is to swap the roles: writing u = sin(x) when the inner function is actually x², which then leaves the integrand unrecognisable. The rubric does not award this row if u does not isolate the inner expression cleanly.
The second row is the differential, du. The student must compute du in terms of x and dx, then solve for dx in terms of du and substitute that into the integrand. The exact form of the differential matters because the constant of proportionality often changes the antiderivative. If the integrand is 2x·cos(x²) and the student writes du = 2x dx, the work is clean. If the student writes du = x dx, the substitution produces a factor of 2 that has to be accounted for later, and a missed factor of 2 is a missed point.
The third row is the antiderivative in u. After substitution, the student writes an antiderivative of the new integrand with respect to u. For a trig composite, this is a standard power rule in u. For an exponential or logarithmic composite, the antiderivative is the inner function unchanged. The rubric awards this row if the antiderivative is correct in u; it does not require the student to translate back to x yet.
The fourth row is the constant of integration, and the fifth row is the back-substitution. These are the two rows that are most often dropped by students who are otherwise competent. A complete antiderivative, after a u-substitution, is F(u(x)) + C, where F is an antiderivative of the rewritten integrand with respect to u. Writing F(u) + C is acceptable, and writing F(x) + C after back-substitution is acceptable; writing F(x) without a constant, or omitting the back-substitution entirely, costs a row.
In my experience, the row that catches candidates is the constant of integration. The graders are looking specifically for a "+ C" somewhere in the work, and a candidate who writes the antiderivative, evaluates it correctly, but never writes "+ C" is leaving a point on the table that is otherwise unmissable. The other row that often slips is the back-substitution, particularly on definite integrals where the student correctly changes the bounds and never needs to substitute back at all — a subtlety that the next section unpacks.
Definite integrals of composite functions: changing bounds vs back-substituting
The single most important distinction in this topic is between an indefinite integral of a composite function and a definite integral of a composite function. The mechanics of the substitution are identical. The bookkeeping is not. In an indefinite integral, the student has two choices for closing out the problem: either evaluate the antiderivative in u and then translate back to x, or evaluate it in u and leave it in terms of u. In a definite integral, the cleaner approach is to convert the bounds from x to u and evaluate entirely in u-space, which removes the need for a back-substitution at all.
The rubric reflects this. On a definite integral FRQ, the row labelled "antiderivative in u" can be followed directly by an "evaluation" row that uses the converted bounds. The "+ C" row, which is mandatory for an indefinite integral, is not required for a definite integral because the constant cancels in the subtraction F(b) − F(a). The grader's checklist for a definite integral looks different: convert u correctly, convert both bounds correctly, evaluate at the upper and lower bound, subtract.
The most common error here is to convert the bounds carelessly. If the original bounds are x = 0 and x = 1, and the substitution is u = x² + 1, then the new bounds are u = 1 and u = 2. A candidate who writes u = 0 and u = 1 — confusing the value of x with the value of x² — has produced a numerically wrong answer that the rubric will mark down. The mechanical translation must be done in the same register: the lower bound stays the lower bound, the upper bound stays the upper bound, only the numbers change.
The second common error is to mix strategies. A candidate converts the bounds, evaluates the integral in u, and then back-substitutes the original x-bounds into the u-expression, producing a term like "(x²+1) evaluated from 0 to 1." This is arithmetically correct in some cases, but the rubric is checking for the substituted bounds specifically. The cleanest scoring path is: pick a strategy, stay with it, and write the bounds explicitly at the moment of evaluation.
For a student preparing for the exam, the tactical advice is to practice both styles. The conversion of bounds is faster and removes one row from the work, but it requires careful arithmetic. The back-substitution approach is slower but more forgiving because the student can verify the answer in the original variable. Either approach is acceptable, and the rubric awards the point either way, provided the steps are written clearly enough to follow.
Question types and where they appear in the AB and BC exams
Composite-function integrals appear in two distinct question types on the AP Calculus exam, and the difference between them drives the preparation strategy. The first type is the stand-alone MCQ: a stem presents an integral, a candidate chooses from four or five antiderivatives, and the work is a single substitution. These items test pattern recognition, not stamina. The second type is the FRQ, where the integral is embedded in a multi-step problem that may also include a differential equation, a particle motion context, or an accumulation function.
On the multiple-choice section, composite integrals are usually short and direct. A typical AB item gives the integrand as a single trig composite, an exponential composite, or a polynomial composite and asks for the antiderivative. The distractors are constructed to penalise specific errors: forgetting the inner derivative, dropping a constant, choosing the wrong sign on a trig function, or confusing the chain rule with the product rule. For these items, the mental process should be: identify the inner function, identify its derivative up to a constant, divide that constant out, integrate the outer function in u, multiply back, and check the answer by differentiation. The whole process should take 90 seconds or less for a well-prepared student.
| Question type | Section | Time budget per item | Key rubric rows |
|---|---|---|---|
| Stand-alone composite antiderivative | MCQ (AB and BC) | ~90 seconds | Correct antiderivative, correct sign, presence of + C only if stated |
| Composite integral inside a context | FRQ (AB and BC) | ~6 minutes of FRQ time | Identification of u, differential, antiderivative in u, +C if indefinite, back-substitution or bound conversion, evaluation |
| Composite integral in u-space with evaluation | FRQ (BC only, often) | ~7 minutes of FRQ time | Bound conversion OR back-substitution, sign in evaluation, correct numerical value |
| Composite integral combined with another technique | FRQ (BC, Unit 8 and 10) | ~8 minutes of FRQ time | Substitution row, then integration by parts or partial fractions row, final numerical answer |
On the free response section, composite-function integrals are almost always embedded in a larger problem. A common AB-style context is an accumulation function: g(x) is defined as the integral from a constant to x of a composite function, and the candidate is asked for g'(x), g″(x), the value of g at a specific point, or the average value of g on an interval. The composite nature of the integrand is the obstacle in the first row, and the candidate has to clear that obstacle cleanly before the rest of the problem becomes accessible.
A common BC-style context is a separable differential equation whose solution requires integrating a composite function. The 2019 BC exam, for instance, asked candidates to solve dy/dx = x·exp(x²) by separation of variables, which requires the same substitution as the 2010 AB FRQ but framed as a differential equation. The rubric treats the substitution as a row and the antiderivative as a separate row, and the candidate has to score both to reach the next stage of the problem.
For a student targeting a 5, the practical preparation move is to drill the stand-alone MCQ pattern until the substitution is automatic, and then to time themselves on FRQ-style problems that embed a composite integral in a 6-minute window. The first 60 seconds should be spent on the substitution itself; the remaining time on the surrounding context. A candidate who cannot identify u in under a minute is going to be pressed for time on the FRQ section, even if the rest of their technique is sound.
Common pitfalls and how to avoid them
Composite-function integration is the topic on which a small number of recurring errors cost the most points. A candidate who has internalised these errors — not just understood them, but rehearsed the correction — will gain a meaningful edge on the exam. The errors cluster into three families: errors of recognition, errors of bookkeeping, and errors of sign.
- Recognition error: missing the inner function. The integrand is 2x·exp(x²) and the candidate sees "exponential," reaches for the formula ∫e^x dx = e^x + C, and writes exp(x²) as the antiderivative. The correct answer is exp(x²) + C, but the candidate has dropped the inner derivative. The fix is to always check: does the integrand contain a function and (a constant multiple of) its derivative? If the answer is yes, substitution is the right move, and the antiderivative is the outer function in u, not in x.
- Bookkeeping error: missing the constant of proportionality. The integrand is x·sin(x²) and the candidate correctly identifies u = x², du = 2x dx, but writes the antiderivative as cos(x²) instead of (1/2)·cos(x²). The factor of 1/2 comes from solving du = 2x dx for dx and substituting dx = du/(2x). The fix is to write the differential row explicitly, including the constant, before integrating.
- Sign error: dropping a negative inner derivative. The integrand is x·cos(x²) and the candidate correctly identifies u = x², du = 2x dx, but writes the antiderivative as sin(x²) instead of (1/2)·sin(x²). This is the same bookkeeping error as above but with a trig function whose derivative is negative — a sign that students often miss because the cosine-to-sine transition is one they are used to seeing in the positive direction. The fix is to differentiate the answer before submitting the work, every time, on every problem.
- Back-substitution error: leaving the answer in u. The integrand is (3x²)·(x³+1)⁴ and the candidate correctly performs the substitution, evaluates in u, and writes (1/5)·u⁵ + C. The full-credit answer is (1/5)·(x³+1)⁵ + C. The fix is to remember that the rubric is checking the variable, not the algebra, and that an answer in u is not an answer in x. The exception, again, is a definite integral with converted bounds, where the u-form is acceptable because the bounds make the variable immaterial.
- Bound error: not converting both bounds. The integrand is x·(x²+1)² from x = 0 to x = 2, and the candidate correctly converts the lower bound to u = 1 but writes the upper bound as x = 2 instead of u = 5. The fix is to substitute the bounds at the moment of conversion, write the converted bounds explicitly, and double-check the arithmetic at the end of the row.
A worked example: a definite integral of a polynomial composite
Consider the integral from 0 to 1 of x²·(x³+1)⁴ dx. This is a standard composite-function problem that has appeared in MCQ and FRQ forms on AP Calculus exams. The integrand contains a polynomial and a polynomial raised to a power, which is the chain-rule shape: the outer function is u⁴, the inner function is u = x³+1, and the derivative of the inner function is 3x². The integrand is x², not 3x², so a factor of 1/3 has to be pulled out before substitution.
The first row is the substitution itself: let u = x³+1, so du = 3x² dx, and therefore x² dx = du/3. The integrand becomes u⁴·(du/3), which is (1/3)·u⁴ du. The antiderivative in u is (1/3)·(u⁵/5) = u⁵/15. The bounds convert as follows: when x = 0, u = 0³+1 = 1. When x = 1, u = 1³+1 = 2.
Evaluating the definite integral in u: u⁵/15 evaluated from 1 to 2 is (2⁵/15) − (1⁵/15) = (32/15) − (1/15) = 31/15. The candidate can leave the answer as 31/15, and the rubric awards the final point for the correct numerical value. No back-substitution is required because the bounds were converted, and no constant of integration is required because this is a definite integral.
For a student preparing for the exam, this example is worth working twice. The first time, work the problem with back-substitution rather than bound conversion: leave the antiderivative in x as (x³+1)⁵/15 and evaluate at x = 1 and x = 0. The answer should be the same 31/15. The second time, work the problem with bound conversion as above. The point of the exercise is to see that both strategies produce the same answer, and to internalise which row of work corresponds to which step in the rubric.
A second worked example: an exponential composite in differential-equation form
For BC candidates, a more representative problem is a separable differential equation whose solution requires integrating an exponential composite. The differential equation dy/dx = x·exp(x²) is a clean example, and the same problem, framed as an FRQ, has appeared in past exam administrations. The task is to find the particular solution satisfying y(0) = 2, which requires the candidate to perform a u-substitution, evaluate a definite integral, and use the result to write the explicit form of y as a function of x.
The first row is separation of variables: dy = x·exp(x²) dx. The second row is integration: ∫dy = ∫x·exp(x²) dx. The third row is the substitution: let u = x², du = 2x dx, so x dx = du/2. The integral becomes ∫exp(u)·(du/2) = (1/2)·exp(u) + C. The fourth row is back-substitution: (1/2)·exp(x²) + C.
The fifth row is the application of the initial condition. Substituting x = 0 and y = 2: 2 = (1/2)·exp(0) + C, so 2 = 1/2 + C, which gives C = 3/2. The particular solution is y = (1/2)·exp(x²) + 3/2.
The rubric for this problem has roughly five rows, of which two are direct scoring of the substitution itself. A candidate who does not perform the substitution — who writes the antiderivative as (1/2)·exp(x²) by pattern recognition alone, without the supporting work — risks losing the row that audits the differential. In my experience, this is the row that separates a 4 from a 5 on this style of problem. The graders are not interested in whether the candidate can produce the right answer; they are interested in whether the candidate can produce a defensible record of the work.
Preparation strategy: drilling the topic in the final weeks
For a student targeting a 5 on the AP Calculus exam, the preparation for composite-function integrals should be spread across the final six weeks of the syllabus, with three distinct phases. The first phase is recognition: working through 30 to 40 stand-alone MCQ items and identifying the inner function and the constant of proportionality in each. The second phase is mechanics: working through 15 to 20 free response problems and writing out the full solution, including the differential row, the antiderivative row, the constant of integration, and the back-substitution. The third phase is integration: working through 5 to 8 mixed FRQs in which a composite-function integral is one step in a larger problem, and timing the work to a 6-minute window.
The first phase should take 4 to 5 hours of focused practice. The second phase should take 6 to 8 hours, with each FRQ worked in full and then graded against the published rubric. The third phase should take 4 to 5 hours, with the candidate timing each FRQ strictly and reviewing the timed solutions against the rubric to identify where time was lost. Across the three phases, the candidate will have written 50 to 70 worked solutions, which is enough to build the muscle memory required to execute a substitution cleanly under exam conditions.
For a student targeting a 4 rather than a 5, the same three-phase plan applies, but the third phase can be shortened and the first phase prioritised. The recognition step is the most time-efficient: a candidate who can identify u in under 30 seconds on a stand-alone problem will rarely struggle on a stand-alone problem at all, and the FRQ items become a question of stamina and rubric discipline rather than conceptual access. A 4-targeted student can reasonably skip the third phase if their time is constrained, and rely on the FRQ practice from their coursework to handle the embedded problems.
For a student who has plateaued, the diagnostic question is always the same: where is the time going? If the candidate can identify u in under 30 seconds but loses points on the differential row, the issue is arithmetic and the fix is to drill the du/dx computation. If the candidate loses points on the back-substitution, the issue is carelessness and the fix is to write the substitution twice — once on the way in, once on the way out — and to verify the final answer by differentiation. If the candidate loses points on the constant of integration, the issue is habit and the fix is to write "+ C" on every indefinite integral, every time, until it becomes automatic.
Conclusion and next steps
Integrals of composite functions are the most heavily weighted single technique on the AP Calculus exam, and the topic where rubric discipline matters as much as conceptual understanding. A student who has mastered the substitution, the differential, the antiderivative in u, the constant of integration, and the back-substitution will score cleanly on roughly 25 percent of the AB free response section and on a comparable slice of the BC exam. The remaining work is to integrate this technique into the larger context of accumulation functions, separable differential equations, and the other FRQ question types, and to time the work to a 6-minute window per question.
AP Courses' AP Calculus preparation programme pairs each student with a tutor who has scored past FRQ administrations against the published rubric and can identify the specific row on which a student is losing points. The first diagnostic session focuses on composite-function integrals, walks through a worked example in real time, and produces a personalised practice plan for the six weeks before the exam.