AP Calculus integration using completing the square is the move that unlocks a surprising number of inverse-trig and logarithm problems on both the AB and BC free-response sections. When a definite integral arrives with a quadratic denominator or a quadratic inside a function, the question is rarely asking you to perform a u-substitution first; it is asking you to recognise that the only way to reach an arctangent, an arcsine, or a logarithm-of-absolute-value answer is to complete the square, separate the integrand, and then integrate each piece with a clean substitution. For candidates aiming at a 5, this algebraic rewrite is the difference between a calculator-free hand-derivation and a frozen first line.
What "completing the square" actually changes inside an AP Calculus integral
Completing the square is a routine algebra move, but inside an AP Calculus problem it does something very specific: it changes the shape of the differential. A quadratic such as x² + 6x + 13 is not friendly to a direct u-substitution, because its derivative 2x + 6 only appears in pieces. Rewriting the same quadratic as (x + 3)² + 4 produces a clean inner function whose derivative shows up exactly once in the numerator (or exactly in the right ratio), which is what an inverse-trig antiderivative needs. The exam does not give credit for memorising the formula; it gives credit for the final antiderivative written on the response line, and the rubric is engineered so that completing the square is the only realistic way to reach the form the scorer expects.
Candidates often underestimate how much of the score is consumed before any calculus appears. On a typical BC free-response item that asks for the area between curves or the volume of a solid of revolution, the algebraic rewrite can take up a full minute of the six minutes the College Board budgets for the problem. In my experience, students who skip the rewrite and try to force a u-substitution on the original expression usually write a wrong antiderivative, then spend another minute trying to fix it, and finish the row with a sign error or a missing +C. The disciplined path is to pause, complete the square, and let the integral become obvious. The rubric never penalises you for spending time on algebra if the antiderivative row is correct.
Two integrand shapes dominate the AP Calculus syllabus. The first is the rational function whose denominator is an irreducible quadratic, which produces an arctangent after completing the square. The second is the form ∫ 1/√(quadratic) dx, which produces an arcsine or a logarithm-of-absolute-value after a slightly different rewrite. Recognising the shape of the denominator before you start writing is the single highest-leverage habit you can build for this topic.
The two AP Calculus integral shapes that require the rewrite
The first shape is the irreducible-quadratic-in-the-denominator integral. A typical stem asks you to evaluate something like ∫ 1/(x² + 4x + 13) dx, often between two values, often as part of a larger accumulation question. The discriminant is negative, the denominator never zero, and the standard arctangent formula cannot be applied directly because the coefficient of x² is not 1 in the right form. The fix is mechanical: take half the linear coefficient (here 2), square it (4), add and subtract, and arrive at (x + 2)² + 9. The integral becomes ∫ 1/((x + 2)² + 9) dx, which is the arctangent formula with a 3 in the denominator of the constant. The general form you should write in your head is 1/(a² + u²) → (1/a) arctan(u/a).
The second shape is the square-root-of-quadratic integral. Stems like ∫ 1/√(-x² + 6x - 5) dx or ∫ 1/√(2x - x²) dx test whether you can recognise the arcsine and the logarithm-of-absolute-value candidates. Completing the square here is more delicate: the quadratic must be rewritten as a constant minus a perfect square, or a perfect square minus a constant, depending on the sign of the leading coefficient. A negative leading coefficient means a constant-minus-square form, which targets arcsine. A positive leading coefficient means a square-minus-constant form, which targets the logarithm of absolute value. Misreading the sign is the single most common error in this whole sub-topic, and the rubric makes no allowance for it.
AP Calculus AB covers the arctangent and logarithm of absolute value forms in Unit 8. AP Calculus BC adds the arcsine form through the integration of inverse-trig derivatives and a handful of logistic and Gaussian pieces in Units 6 and 8. If you are working on the BC side, expect an arcsine to appear in a context where the bounds come from a graph or table, not from a clean algebraic intersection; the AP exam tests whether you can read the bound, not just compute it.
A worked BC example: definite integral with an irreducible denominator
Consider the BC-style problem: evaluate ∫₀² 1/(x² + 4x + 13) dx. The denominator's discriminant is 16 - 52, negative, so the quadratic is irreducible. Completing the square gives (x + 2)² + 9. The bounds stay in x, but you substitute u = x + 2, so the lower bound becomes 2 and the upper bound becomes 4. The integral is now ∫₂⁴ 1/(u² + 9) du, which is (1/3) arctan(u/3) evaluated from 2 to 4. The final answer is (1/3)[arctan(4/3) - arctan(2/3)]. Many candidates lose the row here by writing the antiderivative as arctan((x+2)/3) without the 1/3, because they confuse the formula 1/(a² + u²) with 1/(u² + a²). The constants a and u are not symmetric; the a is what multiplies the arctan on the outside.
How the AP Calculus rubric scores the completing-the-square row
The free-response rubric on the AP Calculus exam is line-based, and the line that matters most for this technique is the antiderivative row. The scorer wants to see, in order, the rewritten integrand, the substitution, the new bounds if it is a definite integral, and the final evaluated expression. Each of those lines has its own point. For a problem worth four points, the breakdown is usually one point for the rewrite, one point for the antiderivative, one point for the evaluation step, and one point for the correct final value. Drop the rewrite and you may still earn the antiderivative point if your form is correct, but you will have spent time re-deriving the form from scratch; the rewrite is a time-saving move as much as a scoring move.
The constant-of-integration row is forgiving in this context. If the antiderivative is an arctangent, no +C is required; the exam treats inverse-trig results as definite-style by convention. If the antiderivative is a logarithm of absolute value, the absolute value is required and missing bars lose the point. If the antiderivative is an arcsine, the function must be written inside the arcsine with the right coefficient on the argument; an arccos or arctan written in its place costs the whole row.
The evaluation row is where the boundaries change. For a definite integral that asks for an exact value, the scorer expects the answer in terms of arctangent, arcsine, or natural log, not a decimal. The exam does allow a calculator-produced decimal for some MCQ items and for selected FRQ parts where a decimal is requested, but the line where the technique is tested is always exact. Candidates who punch a decimal into the calculator on a non-calculator row lose that row outright. The rule of thumb: if the problem does not say "calculator active" and you are dealing with an arctangent, give the exact value.
Where completing the square actually shows up on the AP Calculus exam
The MCQ section of the AP Calculus exam tests the technique less often than the FRQ section, but the MCQ items are the place to build speed. The most common MCQ stem is a definite integral of 1/(ax² + bx + c) where a, b, c are integers and the quadratic is irreducible; the answer choices are always exact values involving arctangent, sometimes in degrees, sometimes in radians. Read the directions for the section you are in. The non-calculator MCQ in Section I, Part A, gives answers in radians; the calculator MCQ in Section I, Part B, sometimes gives both, and you should pick the form the problem asks for. A common trap is an answer choice that looks like the correct arctan form but is missing the 1/a factor from the formula.
The FRQ section, by contrast, embeds the technique in a multi-step problem. A typical AB FRQ on Unit 8 will give you a function defined by a definite integral and ask you to evaluate it at a specific value, find its derivative, or use it as the integrand for a separate area problem. A typical BC FRQ on the same topic will combine the technique with a logistic differential equation, an improper integral, or a separation-of-variables problem. The completing-the-square step is one move in a longer chain, and the rubric gives credit for it as such. Candidates who memorise the technique but do not understand how it sits inside a larger problem often earn the rewrite point and lose the next point because they never connected the rewrite to the next part.
Completing the square versus u-substitution on AP Calculus FRQs
The most common tactical mistake I see is reaching for u-substitution first. A problem that gives you 1/(x² + 4x + 13) does not respond to a u-substitution; the derivative of x² + 4x + 13 is 2x + 4, which is not present in the numerator. If you set u = x² + 4x + 13, you get du = (2x + 4) dx, and you cannot replace any part of the original dx with that. The integral is not in u-substitution form. The same is true of 1/√(x² + 4x + 13). The correct move is to complete the square, then u-substitute inside the rewritten integral.
A second common mistake is the order of operations. The AP Calculus exam is built so that the completing-the-square step is the only way to set up a clean u-substitution. If you start with a u-substitution, you will end up with a du that has an x in it, and the only way to clean that up is to express x in terms of u, which is exactly the algebra completing the square would have given you for free. The two moves are not in competition; they are sequential. Treat the rewrite as a precondition, not an alternative.
There is a third, less common scenario where completing the square is the wrong move. If the integrand is something like x/(x² + 4x + 13), a u-substitution on the denominator is the natural first step, because the numerator is exactly half the derivative of the denominator. In that case, completing the square would only complicate the situation. The rule of thumb: if the numerator is a constant multiple of the derivative of the denominator, u-substitute. If the numerator is a constant, complete the square.
Step-by-step preparation plan for the completing-the-square integral
A solid six-week preparation plan for this technique is straightforward. In the first two weeks, isolate the technique and drill it on ten to fifteen non-calculator problems. The College Board's AP Classroom has tagged problem sets for Unit 8; use the items whose integral stems are irreducible quadratics or square roots of quadratics, and write the full solution by hand. Time yourself: a clean solution should take three to four minutes. If you are taking longer, the bottleneck is the rewrite, not the calculus, and you need to drill the algebra.
In weeks three and four, move to FRQ-style problems. Pick three released FRQs from the BC exam that include an inverse-trig or logarithm antiderivative. The 2012 BC Form B, 2014 BC, and the 2017 BC all have problems that fit. Write out the rubric line by line as you solve, and check your work against the official scoring guidelines. The rubric language for the rewrite row is consistent: the scorer wants to see the completed-square form, the substituted integrand, and the new bounds. If any of those three pieces is missing, that line is gone.
In weeks five and six, fold the technique into mixed-topic timed sets. Pull twenty MCQ items from the released BC exams, half of which include a completing-the-square integral and half of which test other Unit 8 ideas, and give yourself thirty-five minutes. Review the items you got wrong and classify each error: rewrite error, sign error, bounds error, formula error. The classification is the input to the next round of targeted practice. The exam does not reward general studying; it rewards specific, classified error patterns.
Common pitfalls and how to avoid them on completing-the-square FRQs
The first pitfall is the sign of the leading coefficient. For ∫ 1/√(2x - x²) dx, the quadratic 2x - x² rewrites as 1 - (x - 1)², which is the constant-minus-square form. For ∫ 1/√(x² - 2x) dx, the same algebra gives (x - 1)² - 1, the square-minus-constant form. The two forms give an arcsine in the first case and a logarithm of absolute value in the second. Misreading the sign sends you to the wrong formula and costs the whole antiderivative row. The fix is mechanical: factor out the leading coefficient of x² first, complete the square on the inside, then read the form.
The second pitfall is the 1/a factor. The arctangent formula is ∫ 1/(u² + a²) du = (1/a) arctan(u/a) + C. Many candidates write arctan(u/a) and skip the 1/a. On a 0-to-3 definite integral, the numerical difference is small, but on a 2-to-5 integral it shifts the answer, and the rubric reads the antiderivative line literally. A useful habit: write the a value above the integral before you start, and check that the antiderivative is exactly the formula 1/a times the arctan.
The third pitfall is forgetting to update the bounds after substitution. A common stem is ∫₀³ 1/((x-1)² + 4) dx. The natural substitution is u = x - 1, which shifts the bounds to -1 and 2. Candidates who evaluate the antiderivative at 0 and 3 instead of -1 and 2 lose the evaluation row, even though the antiderivative line is correct. The exam does not give a "you obviously meant the right thing" point; the bound row is its own line.
The fourth pitfall is treating the calculator as a backstop. On the calculator-active section, a candidate can compute a definite integral numerically and back-solve, but the rubric for the antiderivative row still requires the symbolic form. The calculator will give you a decimal that matches the correct answer, but the scorer reads the lines above the decimal. If those lines are missing, the row is missing.
Comparison: how AB and BC weight completing-the-square integration
AB and BC treat this technique differently, and the difference matters for time budgeting. The table below summarises the typical distribution of completing-the-square items across the two exams. Use it as a guide for how much of your own practice time to spend on the topic.
| Dimension | AP Calculus AB | AP Calculus BC |
|---|---|---|
| Unit coverage | Unit 8 (Basic Integration) | Units 6, 8, 10 (Logistic ODE, integration, series) |
| Typical MCQ item count | 1 to 2 per MCQ section | 2 to 3 per MCQ section |
| Typical FRQ item count | One part of a single FRQ | One to two parts across two FRQs |
| Antiderig forms tested | Arctan, log of absolute value | Arctan, arcsine, log of absolute value |
| Calculator role | Numerical check only on Part B | Numerical check + logistic ODE work |
| Common pairing | Area between curves | Separation of variables, improper integral |
The takeaway from the table is that BC candidates should expect the technique in a wider variety of contexts, and AB candidates should expect it in a smaller, more contained set. The total number of points at stake is similar across the two exams, but the BC exam distributes them across more problems, which means a single missing row costs a smaller fraction of the total score on BC. For both exams, however, the rewrite row is a guaranteed point on the items where it appears; failing to earn it is almost always an algebraic, not a calculus, error.
Practising the technique under AP Calculus exam conditions
Once the basic drill is comfortable, the next step is to practise under timed conditions that mimic the AP Calculus exam. The non-calculator MCQ section gives you about a minute and twenty seconds per item, and the calculator MCQ section gives you about a minute and thirty-five seconds. The FRQ section gives you roughly fifteen minutes per six-line problem. None of these windows is generous, and the only way to be fast enough is to make the rewrite automatic. A candidate who has to think about completing the square for thirty seconds is using up a quarter of the item's time on algebra; a candidate who has the rewrite in muscle memory spends five seconds on it and has time to check the bounds and the sign.
The most efficient practice is mixed-topic timed sets rather than topic-blocked drills. Topic-blocked drills build accuracy; mixed-topic sets build selection, which is the harder skill on the actual exam. The AP Calculus MCQ section will not flag the items that need completing the square; you have to recognise them on sight. The way to train that recognition is to do twenty MCQ items in a row, half of which are completing-the-square and half of which are other Unit 8 ideas, and time yourself. The classification habit that worked for the long-term preparation plan works here too: every error becomes a data point in your error log.
A final tactical note: keep a one-page reference of the three integral forms, written in your own handwriting, and review it the morning of the exam. The three forms are ∫ 1/(u² + a²) du = (1/a) arctan(u/a) + C; ∫ 1/√(a² - u²) du = arcsin(u/a) + C; and ∫ 1/(u² - a²) du = (1/(2a)) ln |(u - a)/(u + a)| + C. The page should also show a half-worked example for each, including the completed-square form. The act of writing the page by hand, rather than printing it, cements the form in your memory. On exam day, the form should be something you see in your head, not something you reconstruct on the page.
Conclusion and next steps for AP Calculus candidates
Completing the square is a small algebraic move with an outsized effect on the AP Calculus free-response score. It unlocks the irreducible-quadratic denominator that produces arctangent answers, the constant-minus-square form that produces arcsine answers, and the square-minus-constant form that produces logarithm-of-absolute-value answers. The exam rewards the rewrite explicitly, and the rubric reads the antiderivative line literally. Candidates who drill the rewrite, classify their errors, and time themselves on mixed-topic sets will see this technique shift from a stuck item to a guaranteed row.
AP Courses' one-to-one AP Calculus BC programme walks each student through a personal error log on completing-the-square FRQs, classifies every missed row against the official rubric language, and turns the BC Unit 8 inverse-trig integration items into a concrete score-targeting plan for the May sitting.