AP Calculus harmonic and p-series

The AP Calculus harmonic and p-series problem family sits inside the BC-only Unit 10 on infinite series, and it is one of the few places on the exam where a single convergence test decides the answer. Students who walk into the test room uncertain about when to use the integral test, the direct comparison, or the limit comparison routinely throw away points on multiple-choice items that are designed to be solved in under two minutes. This walkthrough builds the framework from the ground up: what a p-series is, why the harmonic series is its boundary case, how the integral test works on these families, and how the BC FRQ rewards candidates who can chain one test to the next.

Where the harmonic and p-series live in the AP Calculus BC syllabus

Unit 10 of the AP Calculus BC course description dedicates roughly 8–10% of the overall exam weighting to infinite sequences and series. Within that unit, the harmonic series and the more general p-series function as the foundational counter-examples that every other convergence test is built around. If a candidate cannot tell at a glance whether the series 1/n, 1/n^2, 1/sqrt(n), or 1/(n^2 + 1) converges, the rest of Unit 10 — comparison tests, ratio test, alternating series, radius of convergence, Taylor polynomials — becomes much harder to reason about.

On the multiple-choice section, harmonic and p-series questions usually appear in the early-to-middle band of Unit 10 items, where the difficulty is calibrated to reward pattern recognition rather than algebraic gymnastics. A typical item asks the candidate to identify which of four series converges, where two are harmonic-like, one is geometric, and one is a p-series with p slightly greater than 1. A 90-second solve is realistic once the student has internalised the p > 1 rule.

On the free-response section, the harmonic and p-series are almost never the final answer. Instead, they are scaffolding: a BC FRQ will ask the student to determine the convergence of a series that looks unfamiliar, but the first move is to compare it to a known p-series. Candidates who skip the comparison step and reach for the ratio test immediately lose 1–2 points per FRQ simply on classification. In my experience grading practice FRQs, the most common error in this family is using the wrong test on the wrong series, not making an arithmetic mistake once the right test is chosen.

Defining the harmonic and p-series precisely

The harmonic series is the infinite sum of the reciprocals of the positive integers: sum from n=1 to infinity of 1/n. It is the canonical divergent p-series, sitting exactly on the boundary between convergence and divergence. The p-series is its generalisation: sum from n=1 to infinity of 1/n^p, where p is a real number.

Three behaviours matter for the AP Calculus exam:

If p > 1, the p-series converges. The most common exam-relevant case is p = 2, the Basel-style sum 1/n^2, which converges to π^2/6, though the candidate is never expected to know that closed form.
If p ≤ 1, the p-series diverges. The harmonic case p = 1 diverges, as do p = 0 (the constant series 1 + 1 + 1 + …) and p < 0 (where the terms do not even tend to zero).
If p is not a positive constant — for example p = n in 1/n^n, or p varies with n — the term is not a p-series, and the student must reach for a different test.

For most candidates, the p > 1 rule is the only thing that needs to be memorised. The test does not require a proof of why p = 1 diverges, only the recognition that it does. That said, the integral test provides the cleanest justification, and writing it out on an FRQ shows the grader that the candidate understands the underlying reason, which can rescue a borderline score.

Why the harmonic series is the boundary case

The harmonic series is the p-series at p = 1, and it is the slowest possible divergence among positive-power p-series. The partial sums grow like ln(n) — the natural logarithm of the number of terms — which means they diverge, but extraordinarily slowly. After 10^6 terms, the partial sum is only about 14. After 10^12 terms, it is about 28. This is why the harmonic series is a perennial favourite for trick questions on the AP exam: a series can look like it is converging because the first few thousand terms behave well, but the divergence is real.

On the AP Calculus exam, the harmonic series shows up most often as a comparison target. A problem will give a series whose terms are slightly smaller than 1/n — for example 1/(n+5) or 1/(2n+1) — and ask whether it converges. The candidate who recognises that 1/(n+5) is asymptotically equivalent to 1/n and applies the limit comparison test will see the divergence in about 30 seconds. The candidate who tries to apply the ratio test will get an inconclusive limit of 1 and waste two minutes before falling back to direct comparison.

A second place the harmonic series appears is in alternating form: sum of (-1)^(n+1)/n. This is the alternating harmonic series, which converges conditionally by the alternating series test, even though the absolute value series (the ordinary harmonic series) diverges. AP candidates should expect at least one item per exam cycle that tests the distinction between absolute and conditional convergence using exactly this series. The bookkeeping is simple: the alternating test gives convergence, but the absolute value series is harmonic, which diverges, so the original series converges conditionally rather than absolutely.

The integral test applied to p-series

The integral test is the most direct convergence test for p-series, and the one the AP Calculus exam expects candidates to be able to write out cleanly. The test states that for a series sum a_n with a_n = f(n) where f is continuous, positive, and decreasing on [1, infinity), the series converges if and only if the improper integral from 1 to infinity of f(x) dx converges.

For the p-series with p ≠ 1, the integral is straightforward:

Integral of 1/x^p from 1 to infinity = integral of x^(-p) dx = [x^(-p+1) / (-p+1)] from 1 to infinity.
If p > 1, then -p+1 < 0, and the limit as x → infinity of x^(-p+1) is 0, giving a finite value of 1/(p-1).
If p < 1, then -p+1 > 0, and the limit is infinity, giving divergence.

For p = 1, the integral of 1/x from 1 to infinity is ln(x) evaluated from 1 to infinity, which diverges. The harmonic series is the case where the integral test produces a logarithm, and that logarithm is the most direct way to see why the harmonic series itself diverges.

On a BC FRQ, the integral test is a high-value move. A question might present the series sum of 1/(n ln n) — the classic borderline series — and ask whether it converges. The integral test gives the integral of 1/(x ln x) dx, which is ln(ln x) from some lower bound to infinity, which diverges. The candidate who sets up the substitution u = ln x, du = dx/x, and reduces the integral to 1/u is showing the grader a full understanding of the test. The candidate who just writes "diverges by the integral test" without setup gets partial credit at best.

Direct and limit comparison tests on p-series

The comparison tests are where the harmonic and p-series do most of their silent work on the AP Calculus exam. The direct comparison test says that if 0 ≤ b_n ≤ a_n and sum a_n converges, then sum b_n converges; and if 0 ≤ a_n ≤ b_n and sum a_n diverges, then sum b_n diverges. The limit comparison test says that if a_n and b_n are positive and the limit of a_n/b_n exists and is a finite positive number, then the two series share a convergence behaviour.

For a typical exam item, the candidate sees a series like sum of 1/(n^2 + n) or sum of n/(n^3 + 1) and must decide. The cleanest move is to compare against a p-series:

Series	Comparison target	Result
sum 1/(n^2 + n)	1/n^2 (since 1/(n^2+n) ≤ 1/n^2)	Converges (p = 2)
sum n/(n^3 + 1)	1/n^2 (asymptotically equivalent)	Converges (p = 2)
sum 1/(2n + 1)	1/n (asymptotically equivalent)	Diverges (harmonic)
sum 1/sqrt(n + 4)	1/sqrt(n), i.e. 1/n^(1/2)	Diverges (p = 1/2)
sum 1/(n^1.5 + n^0.5)	1/n^1.5 (since denominator ≥ n^1.5)	Converges (p = 1.5)

The pattern is mechanical: identify the dominant power of n in the denominator, treat everything else as a bounded constant, and read off the effective p. For most candidates reading this, the key habit is to write down the comparison explicitly on the page. FRQ graders cannot award points for reasoning that does not appear on the page, and "the series converges because the terms go to zero" is not an acceptable justification — the harmonic series is itself a counter-example to that reasoning.

Common pitfalls and how to avoid them

Three errors account for the majority of lost points on harmonic and p-series questions.

First, confusing the term test with a convergence test. The fact that the terms of a series tend to zero is necessary for convergence but not sufficient — the harmonic series is the textbook counter-example, and a candidate who writes "1/n → 0, so the series converges" is making the most-penalised error in the unit. The fix is to internalise the harmonic series as the canonical divergence case where the term test fails.

Second, misreading the exponent. A series like sum 1/sqrt(n) is a p-series with p = 1/2, not p = 2. The square root in the denominator means the exponent is 0.5, and since 0.5 ≤ 1, the series diverges. Candidates who automatically reach for "1/n^2 converges" without rewriting the term in the form 1/n^p lose 1–2 points per item. The fix is to take three seconds and write 1/sqrt(n) = 1/n^(1/2) on the page before deciding.

Third, applying the ratio test where a comparison is faster. The ratio test on a p-series always returns a limit of 1, which is inconclusive. The exam writers know this, and they put p-series-ratio-test combinations in the multiple-choice section specifically to test whether candidates will waste four minutes on an inconclusive test. The fix is a personal rule: if the term is a rational function of n with a single dominant power, reach for comparison first, ratio last.

How harmonic and p-series appear on the BC FRQ

The BC free-response section allots roughly 45 minutes to six questions, and at least one of them is a series question that uses p-series as scaffolding. The structure is consistent: present an unfamiliar series, ask whether it converges, and the candidate's job is to identify the right test and apply it.

A representative 2018-style BC FRQ item might read: "Determine whether the series sum from n=2 to infinity of 1/(n(ln n)^2) converges. Show the work that justifies your conclusion." The candidate who recognises the (ln n)^2 in the denominator should immediately think of the p-test in disguise, or set up the integral test with the substitution u = ln n. The integral reduces to 1/u^2, which converges. The candidate who reaches for the ratio test will spend 90 seconds computing a limit of 1 and then be stuck.

A second representative structure asks the candidate to find the interval of convergence of a power series, and within that interval, the endpoints often produce p-series. For example, the series sum x^n / n converges for -1 ≤ x < 1, with conditional convergence at x = -1 (the alternating harmonic series) and divergence at x = 1 (the ordinary harmonic series). The candidate must classify both endpoints, and the only tool is the alternating series test plus the harmonic divergence fact. For most candidates preparing for the BC exam, this is the single highest-leverage item family to drill.

Scoring on this type of FRQ rewards explicit justifications. A typical 4-point series FRQ allocates 1 point for setting up the test, 1 point for computing the limit or integral, 1 point for stating the conclusion, and 1 point for a correct final answer with proper notation. A candidate who writes the right answer with no work gets 1 point; a candidate who writes the right work and the wrong answer gets 2–3 points. The takeaway: on the BC FRQ, the work on the page matters as much as the final letter.

Preparation strategy and pacing for Unit 10

For a candidate aiming for a 5 on the BC exam, the unit-by-unit weighting suggests spending roughly 15–20 hours of focused study on Unit 10 over a 6–8 week preparation window. Within that time, the harmonic and p-series material deserves a disproportionate share — about a third of the total Unit 10 hours — because it is reused in every later series topic. A candidate who is shaky on p > 1 will be shaky on the comparison tests, the ratio test on borderline cases, and the endpoint analysis of power series.

The most efficient study sequence is: (1) drill the p > 1 rule on at least 30 series classification items; (2) write out the integral test derivation for the harmonic and p-series from scratch, by hand, twice; (3) practice limit comparison on at least 20 series where the comparison target is a p-series; (4) work three full BC FRQ series questions under timed conditions. Candidates who skip step 2 and jump to multiple-choice drilling tend to plateau at a 3 on the AP score scale because they cannot justify their work on the FRQ.

Pacing on the exam itself is straightforward. Multiple-choice items on harmonic and p-series should take 60–90 seconds each. A candidate who spends more than two minutes on a single item has either misread the exponent or applied the wrong test, and the right move is to mark the item, skip it, and return after the rest of the section. On the FRQ, a single series subpart should take 4–6 minutes. Most candidates reading this will find that the bottleneck is not the test selection but the algebraic setup, and that bottleneck is removed by writing the comparison or the integral explicitly before computing.

Putting it all together: a worked example

Consider the series sum from n=1 to infinity of (n + 3) / (n^3 + 2n^2 + 5). The candidate's job is to determine convergence.

Step 1: identify the dominant power. The numerator is linear in n, the denominator is cubic in n. For large n, the term behaves like n / n^3 = 1/n^2. This is a p-series with p = 2.

Step 2: choose a comparison target. The natural target is 1/n^2. Use the limit comparison test: compute the limit of [(n+3)/(n^3+2n^2+5)] / [1/n^2] = (n+3) n^2 / (n^3+2n^2+5) = (n^3 + 3n^2) / (n^3 + 2n^2 + 5). The limit as n → infinity is 1, a finite positive number, so the two series share a convergence behaviour.

Step 3: classify the target. The p-series sum 1/n^2 converges because p = 2 > 1.

Step 4: state the conclusion. The original series converges by the limit comparison test with the convergent p-series sum 1/n^2.

The full written justification should take about three lines on the FRQ. The candidate who writes only the final line — "converges" — leaves 2–3 points on the table. The candidate who writes the four lines above walks away with full credit. In my experience, the difference between a 3 and a 5 on the BC exam is rarely the ability to recognise the right test; it is the discipline of writing the work down so the grader can award the points.

Conclusion and next steps

The harmonic and p-series family is the foundation of AP Calculus BC Unit 10, and a candidate who has internalised the p > 1 rule, the integral test, and the limit comparison test will find the rest of the series unit much more manageable. The exam rewards pattern recognition on the multiple-choice section and explicit justification on the FRQ, so the preparation strategy must drill both. Spend a third of the Unit 10 study time on p-series classification, write the integral test derivation by hand at least twice, and work three full BC FRQ series questions under timed conditions before exam day. Candidates who follow this sequence typically move from a 3 to a 4 or 5 on the AP score scale, and the gain shows up most clearly on the endpoint analysis subpart of power-series FRQs. TestPrep Europe's BC calculus diagnostic is a natural starting point for candidates building a sharper preparation plan around the harmonic series, p-series, and the limit comparison test.

Frequently asked questions

What is the difference between the harmonic series and a p-series on the AP Calculus exam?

The harmonic series is sum 1/n, which is the p-series at p = 1 and diverges. A p-series is sum 1/n^p for any real p; it converges if p > 1 and diverges if p ≤ 1. The harmonic series is the boundary case that the AP exam uses as a counter-example whenever it asks candidates to explain why the term test is not sufficient.

Which convergence test should I use first on a p-series-style question?

If the series is already in the form 1/n^p, the p-series test alone is enough. If the series resembles a p-series but is not in that exact form — for example sum 1/(n^2 + n) or sum n/(n^3 + 1) — reach for the limit comparison test against 1/n^2 or whichever dominant-power p-series matches. Save the integral test for cases with logarithms or other non-algebraic terms, and use the ratio test only as a last resort, since it returns the inconclusive limit of 1 on every pure p-series.

Does the harmonic series ever converge on the AP Calculus exam?

The ordinary harmonic series sum 1/n never converges, and that is a fact the candidate is expected to know. The alternating harmonic series sum (-1)^(n+1)/n does converge, but only conditionally: by the alternating series test the original series converges, but the absolute value series (the ordinary harmonic series) diverges, so the convergence is conditional rather than absolute.

How many points is a typical series question worth on the BC free-response section?

A series subpart on the BC FRQ is usually worth 2–4 points out of the 9 points per question. A 4-point item typically allocates 1 point for setting up the correct test, 1–2 points for carrying out the computation or comparison, and 1 point for stating the conclusion with proper notation. Candidates who skip the setup and write only the final answer typically receive 1 point, which is why the written work matters as much as the final letter.

How much of the AP Calculus BC exam covers the harmonic and p-series material?

Unit 10 on infinite sequences and series accounts for roughly 8–10% of the BC exam weighting, and the harmonic and p-series material is the foundation of that unit. While the harmonic and p-series rarely appear as the final answer on an FRQ, they appear as the comparison target or classification reference on the majority of Unit 10 items, so mastery of p > 1 and the harmonic counter-example has an outsized effect on the overall BC score.