When does the integral test actually work on the AP

The integral test for convergence is a tool that lets a student decide whether an infinite series converges or diverges by comparing its terms to a related improper integral. On the AP Calculus BC exam, this idea sits inside Unit 10 of the course framework, the series unit, and it is one of the cleanest ways to turn a series question into a calculus question you can actually compute. The test has three requirements — positive terms, a continuous, positive, decreasing function, and a tractable integral — and if all three hold, the series and the integral share the same fate. Used correctly, it is the workhorse behind about a third of the series MCQ items in the BC Multiple-Choice section, and it shows up as a stepping stone in 1–2 free-response questions each sitting.

Why the integral test earns its place on the AP Calculus BC exam

The AP Calculus BC course framework dedicates a portion of Unit 10 to convergence tests for series of constants. The integral test is the first convergence test students meet that bridges two big ideas: the integral as accumulation, and the series as partial-sum behaviour. From a score-building point of view, this is one of the most efficient items to master, because the mechanics are short, the conceptual payoff is large, and the same template recurs in disguised forms across multiple questions.

On the exam, the integral test is rarely the final answer. It is usually the first move in a chain. A typical BC free-response item walks the student through: (a) decide convergence or divergence of a p-series; (b) apply the integral test to a non-standard series; (c) use the remainder estimate to bound the error of a partial sum. The integral test is the hinge between part (a) and part (b), so losing it costs points twice. In my experience grading mock BC papers, a student who gets the integral test right on one part usually carries that fluency into the comparison and ratio test questions on the next part. Lose it, and the comparison and ratio items that follow become much harder.

For exam preparation, the high-yield return is enormous. A single 30-minute drill on integral test mechanics, conditions, and p-series shortcuts can lift the series sub-score by a full raw point on the multiple-choice section, which translates to a noticeable movement on the 1–5 scale. Students aiming at a 5 in BC should treat the integral test as compulsory, not optional, and should be able to write out the three conditions from memory before the proctor says begin.

The three conditions of the integral test that the AP reader will check

The integral test is not a free pass. The College Board explicitly tests whether you verify the three conditions before you compute the integral, and graders do take points off for skipping them. The conditions, written for a series ∑ aₙ where aₙ = f(n) for n ≥ 1, are these.

First, the function f must be continuous on [1, ∞), or more generally on the interval [N, ∞) for some positive integer N. Discontinuities on the interval can be addressed by starting the series at a later index, which is a common exam trick. A typical BC question gives you a series that begins at n = 2 or n = 3 precisely so the test still applies once you adjust the index.

Second, f(x) must be positive for all x ≥ N. Negative or sign-changing series do not get the integral test. The test compares a sum of positive terms to an area under a positive curve, so the geometry only makes sense if everything is positive. If you see (−1)ⁿ in the series, walk away from the integral test and reach for the alternating series test instead.

Third, f(x) must be decreasing on [N, ∞). Geometrically, this is what allows the rectangles in the Riemann comparison picture to sit either above or below the curve, producing the bounds that drive the test. Algebraically, you verify it by showing f'(x) ≤ 0 on the interval for the differentiable case, or by direct inequality for sequences. The decreasing condition is the one students most often forget to check on a multiple-choice item, and it is the easiest place to lose a point on a free-response.

A useful exam tactic is to write the three conditions in a single bullet list as the first line of your work, then check off each one before moving on. It costs you ten seconds and saves you a one-point deduction when the grader is moving quickly through a stack.

The core formula and how to set it up on the exam

With the three conditions met, the integral test states that the series ∑ f(n) converges if and only if the improper integral ∫₁^∞ f(x) dx converges. The integral and the series share the same convergence behaviour: both converge, or both diverge. Note carefully — the integral test does not give you the value of the series, only the verdict.

The set-up is mechanical. Replace n with x, replace the summation index with the lower limit, and replace infinity with the upper limit. So the series ∑_{n=1}^∞ 1/n² becomes the integral ∫₁^∞ 1/x² dx. The series ∑_{n=2}^∞ 1/(n·ln n) becomes ∫₂^∞ 1/(x·ln x) dx. The substitution is uniform, and once you have made it, you compute the improper integral using a limit of a definite integral as the upper bound goes to infinity.

There are three integral shapes that appear over and over on BC exam items. The p-integral ∫₁^∞ 1/xᵖ dx converges if p > 1 and diverges if p ≤ 1. The integrand 1/xᵖ is the p-series template, and the result gives you the entire p-series classification for free. The exponential decay integral ∫₁^∞ 1/aˣ dx with a > 1 is a geometric-area style computation and always converges. The logarithmic integral ∫₂^∞ 1/(x·(ln x)ᵖ) dx is the only one that needs substitution, and it converges if and only if p > 1. This last one is the BC exam's favourite trap, and it is worth memorising in both the integral and series forms.

Most candidates reading this should know that the integral test rarely requires a tricky antiderivative. If your integral set-up needs integration by parts, a trig substitution, or partial fractions, you have probably chosen the wrong convergence test. Reach for direct comparison or limit comparison instead, which are built for the harder integrals.

Worked example 1: a clean p-series disguise

Consider the series ∑_{n=1}^∞ 1/n³. Verify the integral test applies. f(x) = 1/x³ is continuous on [1, ∞), positive, and decreasing because the derivative −3/x⁴ is negative. The integral ∫₁^∞ 1/x³ dx evaluates to lim_{b→∞} [−1/(2x²)] from 1 to b = 0 − (−1/2) = 1/2. The integral converges, so the series converges. Two lines of work, full credit on a free-response.

Worked example 2: a logarithmic integral trap

Consider ∑_{n=2}^∞ 1/(n·(ln n)²). Here f(x) = 1/(x·(ln x)²) is continuous on [2, ∞) (note the lower bound), positive, and decreasing. The integral ∫₂^∞ 1/(x·(ln x)²) dx. Let u = ln x, du = dx/x. The integral becomes ∫_{ln 2}^∞ 1/u² du, which equals 1/ln 2. The integral converges, so the series converges. This is a classic BC question and the substitution is the only tricky step — if you can do it cold, the rest of the question is free.

How the integral test compares to direct and limit comparison on the BC exam

Convergence tests are a family, and the BC exam is explicit that students should be able to choose between them. The integral test is the most structured: it requires an integrable function and a manageable antiderivative. Direct comparison requires a known benchmark series and a pointwise inequality. Limit comparison requires a known benchmark series and a finite, positive limit of ratios.

A useful rule of thumb I share with my own BC students: if the term contains 1/xᵖ alone, use the integral test. If the term contains a polynomial in n multiplied by a power of 1, use comparison against a p-series. If the term contains an exponential, use ratio test. If the term contains a factorial, use ratio test. If the term contains a logarithm inside a polynomial structure, try integral test first, fall back to comparison if the integral is ugly.

Many BC questions give you a series that looks integrable but actually has a small twist — a missing power, a square root, an extra n in the denominator. The integral test still works in all of these if the integral is reasonable, but the bound on the integral test (the partial sum error estimate) is not always easy to compute. The exam often pairs the integral test with a remainder estimate in the next sub-part, so when in doubt, check whether the resulting integral has a closed form. If it does, the integral test is almost certainly the intended path.

Common pitfalls and how to avoid them on the AP Calculus exam

The integral test is short, but the exam has a long memory for its pitfalls. The first pitfall is forgetting the third condition — that the function must be decreasing. A series like ∑ sin(n)/n looks like it might pass the test, but the function is not positive, and the integral test simply does not apply. The correct response is to use absolute convergence plus the alternating series test, or simply to reach for comparison or limit comparison. The grader is not looking for cleverness; they are looking for the right test for the right series.

The second pitfall is misreading the index. A series that starts at n = 2 forces the integral to start at 2, and a series that starts at n = 0 cannot have the integral test applied at all because the integral test requires a positive integer lower bound. If the problem gives you a series starting at n = 0, rewrite it as the sum of the n = 0 term plus a series starting at n = 1, then apply the test to the tail.

The third pitfall is treating the value of the integral as the value of the series. They are not equal, and a free-response that writes ∑ 1/n² = π²/6 on the basis of the integral test alone will lose a point. The integral test only tells you whether the series converges. For the actual value, you need a different tool — usually a known power series expansion or a Taylor polynomial approximation, which belongs to a different unit of the BC framework.

The fourth pitfall is wasting time on an ugly integral. If your set-up produces an integral you cannot evaluate, abandon the integral test, try comparison, and move on. A common MCQ pattern is to give you a series that is, in principle, integrable, but whose antiderivative involves a non-elementary function. The exam is testing whether you recognise that the integral test is not the right tool. The skill here is to set up the integral, see the trap, and switch.

The fifth pitfall is over-using the integral test. AP Calculus students trained in AB often carry the habit of always trying the integral test, and on BC it costs them. Direct and limit comparison are faster on most polynomial-style series, and the ratio test is faster on any series containing aⁿ, n!, or 1/aⁿ. The integral test is a precision tool — use it where it shines, and reach for something else where it does not.

Remainder estimation: how the integral test pays off in part (c) of free-response items

The BC exam's free-response section routinely includes a sub-part that asks the student to bound the error of a partial sum approximation. The integral test gives a clean way to do this, because the rectangles used in the comparison picture can be read as upper and lower bounds on the partial sum. Specifically, if f is positive and decreasing on [1, ∞), then the remainder R_N = ∑_{n=N+1}^∞ f(n) satisfies

∫_{N+1}^∞ f(x) dx ≤ R_N ≤ ∫_N^∞ f(x) dx.

The exam often phrases this as: how many terms are needed to approximate ∑ f(n) to within 0.01? You solve the integral inequality, find the smallest N for which the upper bound is below the tolerance, and you have your answer. This remainder estimate is a 2–3 point sub-part on most BC FRQs, and it is a direct application of the integral test's geometry.

The trick on the exam is to remember the inequality direction. The upper bound is the integral that starts at N, the lower bound is the integral that starts at N + 1. Students who flip the inequality lose a point even when the rest of the work is correct. A short mnemonic: U for upper and N for N (the smaller index) — they go together.

Practice question types you should drill before the BC exam

There are five integral test question types that appear in the College Bank and the released BC exam papers, and a candidate who can handle all five has covered 80%+ of the integral test scoring opportunity. The first type is the direct p-series classification: a series of the form 1/nᵖ or √n / nᵖ, with the question asking convergence or divergence. The integral test gives the answer in a single line. The second type is the logarithmic integral: a series of the form 1/(n·(ln n)ᵖ) or 1/(n·(ln n)·(ln ln n)ᵖ). These require the substitution u = ln x and a p-integral inside.

The third type is the exponential integral: a series of the form 1/(aⁿ·n) or 1/(n²·2ⁿ). The integral test converges easily because the integral is dominated by a geometric series, but the comparison test is usually faster. The fourth type is the remainder estimate, already covered in the previous section. The fifth type is the chain problem: a series whose convergence is decided by the integral test, whose value is decided by comparison to a known series, and whose partial sum is decided by a Taylor polynomial. The BC exam's harder free-response items are built on this kind of layering, and a strong student should be able to identify which sub-part needs which tool.

For preparation, work through at least ten BC-style integral test questions in the final two weeks of your study plan, with the answer key hidden until you have written the three conditions and the set-up integral for each one. The discipline of writing the conditions first is what carries you through the harder sub-parts on the actual exam.

Study-planning and exam-format strategy for the BC series unit

Unit 10 of the AP Calculus BC framework contains roughly 8–10% of the multiple-choice weighting and a guaranteed free-response question. Within that unit, the integral test is one of six convergence tests, and the relative weight is roughly: integral test 15%, direct comparison 20%, limit comparison 20%, ratio test 20%, alternating series test 15%, and root test 10%. The integral test is therefore not the highest-weight test, but it is the test on which students gain the most points per minute of study, because its conditions are explicit and its integrals are short.

My recommended study schedule is three sessions of 45 minutes each, spaced across a week. Session one: read the conditions, derive the test geometrically, work three p-series examples. Session two: work three logarithmic integral examples, plus the substitution u = ln x. Session three: work three remainder estimate examples, plus one full BC-style free-response sub-chain. By the end of the third session, the integral test should feel mechanical, and the student should be able to recognise the relevant question types within ten seconds of seeing the series.

On exam day, the practical advice is simple. When you see a series on the BC Multiple-Choice section, do not default to the integral test. Read the term. If it is a clean 1/xᵖ shape with a polynomial-like denominator, use the integral test or direct comparison. If it contains an exponential or a factorial, use the ratio test. If it alternates in sign, use the alternating series test. The integral test is a precision tool that pays off in roughly one out of every three series questions, and using it selectively is what separates a 4 from a 5 in the BC series sub-score.

Conclusion and next steps

The integral test for convergence is the cleanest convergence test on the AP Calculus BC exam, and it is the only one whose conditions and applications can be fully memorised in a single focused study session. Verify the three conditions, set up the improper integral, evaluate, and you have your verdict. Use it on p-series and logarithmic-integral series, pair it with a remainder estimate in the next sub-part, and you have covered the majority of the BC series scoring opportunity. Build a habit of writing the three conditions before the integral, and you will not lose the easy points that the graders love to deduct.

TestPrep Europe's BC series diagnostic is built around exactly this sequence — condition check, set-up, evaluation, and remainder estimate — and it is the natural starting point for candidates building a sharper preparation plan for the Unit 10 sub-score.

Frequently asked questions

Does the integral test for convergence appear on the AP Calculus AB exam?

No. Convergence tests for series of constants, including the integral test, are part of the AP Calculus BC framework only. AB students are not assessed on this material, although they may see geometric series and the harmonic series in pre-Unit 10 contexts. BC students should expect at least one integral test item in the multiple-choice section and at least one sub-part in the free-response.

What are the three conditions of the integral test, and what happens if one fails?

The function f(x) must be continuous, positive, and decreasing on [N, ∞) for some positive integer N. If any condition fails, the integral test is not valid for that series and another test — typically direct comparison, limit comparison, ratio test, or alternating series test — should be used. The exam routinely gives a series that fails one condition, and the correct response is to identify the failure and switch tools.

Can the integral test give the actual value of a convergent series?

No. The integral test only decides whether the series converges or diverges; it does not return a numerical value. To find the actual value, students need a different technique, most often a known power series expansion or a Taylor polynomial approximation. Confusing convergence with value is a common exam error.

How does the integral test interact with the remainder estimate on a free-response?

The integral test's geometry produces a natural bound on the partial-sum error: the remainder R_N is sandwiched between ∫_{N+1}^∞ f(x) dx and ∫_N^∞ f(x) dx. The exam's free-response sub-parts often ask the student to find the smallest N for which the upper bound falls below a stated tolerance, typically 0.01 or 0.001. This remainder estimate is worth 1–2 raw points on a typical BC free-response.

When should a student choose direct comparison or limit comparison over the integral test?

Direct and limit comparison are usually faster on polynomial-style series, and they are the right choice when the corresponding integral is not elementary or would require integration by parts. As a rule of thumb, use the integral test for clean 1/xᵖ and 1/(x·(ln x)ᵖ) shapes, and reach for comparison when the term has an extra polynomial factor that would make the antiderivative hard. The exam tests the ability to choose the right test, not just to compute a given one.

When does the integral test actually work on the AP Calculus BC exam?