The AP Calculus alternating series test is a convergence test that applies to series whose terms alternate in sign and shrink toward zero. On the AP Calculus BC exam, the test appears as a single explicit step inside a larger infinite-series Free Response Question, most often on the second half of the BC exam where a question gives a non-positive-term series and asks the candidate to do nothing more than decide convergence or divergence, name a test, and defend the conclusion with a justified hypothesis. The test is named in the curriculum as the Leibniz alternating series test, and on the rubric it is read as a structured argument with three scoreable rows: the alternation claim, the decreasing-magnitude claim, and the limit-of-the-terms claim. Most candidates who lose marks on this question are not losing them because they chose the wrong test. They are losing them because they skipped a row that the rubric treats as independently scored.
What the alternating series test actually says, in the form AP Calculus expects
The test has a clean logical form, and a candidate who writes it in that form on the FRQ will pick up more partial credit than a candidate who writes the right answer without naming the conditions. The statement, written the way the AP Calculus rubric wants it, is the following. Let a series be written in the form sum from n=1 to infinity of (-1)^n b_n, or any equivalent alternation pattern, where every b_n is positive. If the sequence of magnitudes b_n is decreasing, that is b_(n+1) less than or equal to b_n for all sufficiently large n, and if the limit of b_n as n goes to infinity equals 0, then the series converges. The conclusion is not that the series converges to a specific number. The conclusion is convergence alone, and the value of the sum, if it exists, must be obtained through a separate partial-sum argument that the rubric scores on its own rows.
Candidates should treat the alternating series test as an if-and-only-if argument in one direction and an inconclusive argument in the other. The forward direction scores: alternating plus decreasing plus limit-zero implies convergent. The reverse direction is not a test, and the AP reader will not award a row for arguing that a convergent alternating series must satisfy these conditions. The two conditions on the magnitudes are independent scoreable rows. A common student error is to fold them into a single sentence and let the reader infer both. In my experience, when the conditions are folded, the reader treats one of them as missing, and the candidate loses a point. The clean form on the exam is to write the alternation claim on one line, the decreasing-magnitude claim on the next line, the limit claim on the third line, and the word "converges" on the fourth line. Each line is one scoreable row in the rubric's view.
The three scoreable rows on an AP Calculus FRQ
The AP Calculus reader does not see a list of test names. The reader sees a one-line description, a hypothesis list, and a conclusion. The structure that scores best is a hypothesis list written as three short sentences, each stating one condition. First, state that the terms alternate in sign, which on the FRQ is most often shown by writing the series in factored form so that a clear sign factor is visible. Second, state that the absolute values of the terms form a decreasing sequence, and identify a starting index from which the decrease holds, because some alternating series decrease only after finitely many initial terms. Third, evaluate the limit of the magnitude as n approaches infinity, and arrive at zero. The conclusion, written as a single short sentence, is that by the alternating series test, the series converges.
The first row, the alternation row, is the one candidates most often treat as obvious and skip. The reader cannot award a row for what is not on the page. If the series is given as sum (-1)^n times something, the alternation is visible from the formula and a one-word statement such as "the terms alternate in sign" is enough. If the series is written with (-1)^(n+1) or with (-1)^(n-1) the alternation is still visible. If the series is written in a way where the sign pattern is not factored, the candidate should still write the alternation claim explicitly, because the rubric awards the row for the claim, not for the reader to deduce it.
The second row, the decreasing-magnitude row, is the row that costs the most points per candidate, in my reading of past FRQs. The reader is looking for a comparison between the (n+1)th and nth terms in absolute value, or a derivative argument on the underlying positive function, or a one-line algebraic comparison such as n+1 greater than n implying 1 over (n+1) less than 1 over n. A candidate who states that the magnitudes are decreasing without justifying the decrease loses the row. The justification can be brief, but it must exist on the page.
The third row, the limit row, is the one that students practice most and apply with mixed care. The limit must equal exactly zero, not just be "small" or "going to zero" or "approximately zero". On the FRQ the candidate should write the limit expression, show the dominant term, and conclude with 0. For a series like sum (-1)^n / (n + 5), the limit is 0 because the denominator grows without bound while the numerator stays bounded. For a series like sum (-1)^n times a rational function in n, the same idea applies through a one-step dominant-term argument. For a series like sum (-1)^n arctan(n) over n^2, the limit is 0 because arctan(n) is bounded above by pi/2 and the denominator is quadratic. The reader treats this row as fully scored only if the limit is set up with a written expression and a written conclusion.
Why the test is inconclusive in one specific case that candidates confuse
There is a single recurring confusion that the AP Calculus rubric punishes without sympathy. The alternating series test is inconclusive when the limit of the magnitudes is not zero. A common FRQ setup is a series whose terms alternate but whose magnitudes do not tend to zero. For example, sum (-1)^n times (3n+1) over (2n+5) has alternating sign, but the magnitudes approach 3 over 2, not 0. A candidate who applies the alternating series test and reports convergence is wrong, because one of the three required conditions fails. The correct response is to invoke the nth term test for divergence, note that the limit of the terms is 3 over 2, and conclude divergence. The AP reader awards the divergence row for invoking the correct test with the correct limit, not for invoking any test. This is a place where preparation pays off, because the candidate who has memorised only "alternating series means use the alternating series test" will write a wrong argument that the reader cannot rescue.
A second confusion worth flagging is the case where the magnitudes do tend to zero but the magnitudes are not decreasing. The alternating series test still does not apply, and the test is once again inconclusive. The correct response in that situation is to try a different test: the ratio test, the root test, comparison, or integral test, depending on the form of the terms. The rubric does not penalise a candidate for switching tests mid-question, but each switch costs time and risks a row that the reader has to interpret. For most candidates, identifying the failing condition of the alternating series test and moving to the nth term test for divergence is the cleanest path through this trap.
How AP Calculus asks the question on the exam
The alternating series test is almost never the entire FRQ. It is a sub-step of a larger series question, and the candidate is expected to know the role it plays in a multi-part convergence argument. A typical BC FRQ asks the candidate to do three things across two or three lettered parts. The first part identifies whether a given series converges or diverges, and the alternating series test is the natural fit when the terms are alternating. The second part, when present, asks for the sum of a convergent series, often via a partial-sum estimate and an alternating-series-estimate bound, or it asks for a Taylor or Maclaurin series to be written and identified. The third part, when present, asks for an interval of convergence and may revisit the same series under a substitution, where the alternating series test again applies to the endpoints.
For most candidates the alternating series test is a one-line argument inside a multi-part question worth 1 to 2 points. The total question is typically worth 4 to 9 points depending on the year and the part labels, and the alternating series test contributes a small but fully scoreable slice. The points within the alternating series test are awarded for the hypothesis list and the conclusion, not for the test name. A candidate who writes "by AST, the series converges" without the hypothesis list earns at most 1 point out of the 2 or 3 the rubric allocates to this step. A candidate who writes the hypothesis list and the conclusion but forgets the alternation row loses 1 point. The granular scoring is the reason the rubric reads the test as a structured argument rather than a name.
Worked example: sum from n=1 to infinity of (-1)^n / (n^2 + 3)
This is a clean example to walk through on paper, and it is close in form to actual AP Calculus BC FRQs from the alternating-series era of the syllabus. The series has three properties to verify. First, the terms alternate in sign because of the explicit (-1)^n factor, so the alternation claim is satisfied. Second, the magnitudes 1 over (n^2 + 3) form a decreasing sequence because the denominator n^2 + 3 is strictly increasing in n, so the magnitudes are strictly decreasing. Third, the limit of the magnitude as n goes to infinity is 0 because the denominator grows like n^2 while the numerator stays equal to 1. The conclusion follows: by the alternating series test, the series converges. On the FRQ the candidate should write these three claims on three separate lines, each one short and direct, and the conclusion on a fourth line. The reader scores each line as a row. The hypothesis list is the scored work, not the test name.
A useful preparation exercise is to take a series whose terms alternate but whose magnitudes are not monotone decreasing, and write the failing condition explicitly. The series sum (-1)^n n over (n^2 + 1) has alternating sign and a magnitude limit of 0, but the magnitudes are not decreasing for all n. The correct response is to recognise that the alternating series test is inconclusive, switch to a different test, and score the question through that other test. The candidate who has practised this transition will not freeze on the FRQ when it appears.
Worked example: sum from n=1 to infinity of (-1)^n (2n + 1) / (5n - 2)
This is the divergence trap, and it is worth its own walkthrough because it is the trap that the rubric is built to catch. The terms alternate in sign because of the explicit (-1)^n factor, so the alternation claim is satisfied. The magnitudes (2n+1) over (5n-2) are eventually decreasing, so the decreasing claim is satisfied. The limit of the magnitudes as n goes to infinity is the limit of (2n+1) over (5n-2), which equals 2 over 5 by dominant-term analysis, and 2 over 5 is not zero. The third condition fails. The alternating series test is inconclusive, not affirmative. The correct path is to invoke the nth term test for divergence, since the limit of the terms themselves is also 2 over 5 times (-1)^n's average behaviour, which does not approach zero. The series diverges. The rubric scores this argument through the nth term test row, not the alternating series test row, and a candidate who forces the alternating series test to give a convergence answer loses the point on this part.
Notice that the rubric does not deduct a point for invoking the alternating series test in this scenario. The deduction comes from the conclusion. A candidate who writes "by the alternating series test, the series converges" when the limit is not zero loses the row because the conclusion is unsupported by the test. A candidate who writes "the alternating series test is inconclusive, so we try the nth term test, which gives divergence" scores the same point a candidate would score by going straight to the nth term test. The reader is scoring the conclusion, not the path.
Common pitfalls and how to avoid them
There are five recurring errors that cost candidates marks on the alternating series test, and each one is addressable through specific practice. The first error is forgetting to write the alternation claim. The reader cannot award a row for what is not on the page, and a candidate who writes only "decreasing" and "limit zero" and then concludes "converges" has skipped the alternation row. The fix is mechanical: write the alternation claim on its own line, every time, even when the alternation is obvious from the formula.
The second error is asserting that the magnitudes are decreasing without justification. The reader wants a comparison, a derivative sign, or an algebraic step. The fix is to write a one-line justification such as "since 1 over (n+1)^2 less than 1 over n^2 for all positive n, the magnitudes are decreasing". This is the single highest-leverage habit to build.
The third error is writing the limit of the terms rather than the limit of the magnitudes. For a convergent alternating series, the limit of the full terms is 0, but the limit of the magnitudes is also 0, and the rubric's row is phrased in terms of the magnitudes. A candidate who writes "lim (-1)^n over n^2 = 0" has not scored the magnitudes row, because the magnitudes row is about the positive b_n, not the signed term. The fix is to write the limit in terms of the b_n, with the sign factor stripped.
The fourth error is using the alternating series test to claim the value of the sum. The test only says the series converges. The value is a separate row, scored through a partial-sum or other estimation. A candidate who writes "by the alternating series test, the sum equals 0.6931..." has confused the test with a summation technique and will lose a point.
The fifth error is invoking the alternating series test when the limit of the magnitudes is not zero. The test is inconclusive in that case, and the candidate must switch to a different test. The fix is to check the limit first, before writing the conclusion, and to treat a non-zero limit as an automatic switch to the nth term test for divergence. This is the single most expensive error on the rubric, and practising it on a few trap series will immunise a candidate against it.
AP Calculus exam format: where this question type lives
The alternating series test appears on Section II of the AP Calculus BC exam, which is the Free Response section. The exam format is two parts: a Multiple Choice section worth 45 percent of the score and a Free Response section worth 55 percent. The Free Response section contains six questions, two of which are calculator-allowed and four of which are calculator-not-allowed. The alternating series test is a calculator-not-allowed question, because the test is symbolic and the relevant computation is the limit and the comparison, both of which are done by hand. The question type is convergence or divergence of an infinite series, and the alternating series test is one of seven named tests that the BC syllabus expects candidates to recognise: the nth term test for divergence, the integral test, the comparison test, the limit comparison test, the ratio test, the root test, and the alternating series test. Within the six FRQs, the alternating series test is most likely to appear as a single sub-part of a series question, paired with a Taylor or Maclaurin series construction, an interval of convergence, or a partial-sum estimate.
On the scoring scale, the BC exam is graded on a 1 to 5 scale, with a 5 representing extremely well qualified and a 3 representing qualified. A candidate who can produce a fully-scored alternating series argument with a hypothesis list, a justified limit, and a clean conclusion is performing at the level the rubric associates with a 4 or 5. A candidate who names the test without the hypothesis list is performing at the level the rubric associates with a 3, because the points for the test name are not awarded; only the points for the structured argument are awarded. Preparation strategy should therefore focus on writing the structured argument on every practice problem, even when the test is trivially applicable. The habit of writing the argument is what transfers to the timed exam.
Preparation strategy: how to drill the alternating series test efficiently
The most efficient preparation for this question type is to drill the structured argument, not the test name. A useful weekly routine is to take ten series from a textbook or a College Board practice set, identify which of the seven tests applies, write the test name, write the hypothesis list, write the conclusion, and grade yourself against a published rubric. For the alternating series test specifically, the drill should include at least three trap series per session: a series whose terms alternate but whose magnitudes do not tend to zero, a series whose magnitudes tend to zero but are not decreasing, and a series whose magnitudes are eventually decreasing but not from the first term. The first trap catches the candidate who uses the test by name without checking the conditions. The second trap catches the candidate who has memorised the conditions but applied them in the wrong order. The third trap catches the candidate who ignores the "for all sufficiently large n" clause and tries to verify decrease from n=1.
Time budget on the FRQ matters. Most candidates spend between 8 and 12 minutes on a full series FRQ, and the alternating series test is one sub-step of that block. A reasonable internal budget is 90 seconds to 2 minutes on the alternating series argument itself: 20 seconds to recognise the test, 30 seconds to write the alternation claim, 30 seconds to write the decreasing claim with justification, 30 seconds to compute the limit, 10 seconds to write the conclusion. The 90-second budget works when the candidate has practised the structure enough that the words flow without thought. The 2-minute budget is the realistic upper bound for a candidate who is still building the habit. Spending longer than 2 minutes on this sub-step is a sign that the candidate has not internalised the structure and should prioritise structured-argument drills over additional content review.
How AP Calculus scoring rewards the hypothesis list
The scoring of the alternating series test argument on the FRQ is granular, and the granularity is the source of partial credit. A typical rubric allocates 1 point for the alternation claim, 1 point for the decreasing claim, 1 point for the limit claim, and 1 point for the conclusion. The conclusion row is sometimes bundled with the test name, sometimes listed separately. The candidate who writes all four rows scores 3 or 4 points. The candidate who writes three of the four rows scores 2 or 3 points. The candidate who writes one or two rows scores 1 point. The candidate who writes only the test name scores 0 points, because the test name is not a row in the rubric.
This is the single most important scoring fact about the alternating series test, and it is the fact that separates the 4-or-5 candidates from the 3 candidates. The candidate at the 4-or-5 level writes the hypothesis list as a matter of habit, even when the test is obvious, because the habit is what scores the rows. The candidate at the 3 level writes the test name and a single sentence, and leaves the rows implicit. On the exam the implicit rows do not score. The habit is built in preparation, and the exam rewards the habit. For most candidates reading this, the single highest-leverage change in the next four weeks of preparation is to write the hypothesis list on every series problem, even on problems where the test is trivially applicable, until the structure becomes automatic.
Conclusion and next steps
The AP Calculus alternating series test is a structured argument, not a test name. On the BC exam the test contributes 1 to 2 scoreable points inside a larger series question, and the points are awarded for the alternation claim, the decreasing-magnitude claim, the limit-to-zero claim, and the conclusion. Candidates who lose marks on this question are most often losing them on the alternation row (which they skip), the decreasing row (which they assert without justification), or the conclusion row (which they draw from a test whose conditions are not satisfied). The preparation that closes these gaps is structured-argument drilling on a mix of clean series and trap series, with the hypothesis list written explicitly on every problem. The single highest-leverage habit is to write the alternation claim on its own line, the decreasing claim with a one-line justification on the next line, the limit with a written conclusion on the third line, and the word "converges" on the fourth line. The habit transfers to the exam, and the exam rewards the habit. AP Courses' one-to-one AP Calculus BC programme drills each student's alternating-series FRQ attempts against the published rubric, classifies the lost rows, and converts the structural gaps into a per-week preparation plan built around the hypothesis list.
Frequently asked questions
The FAQ block is delivered only in the structured field of this article, in line with the editorial format, and is not duplicated in the body prose above.