A Riemann sum approximates a definite integral by partitioning an interval into subintervals and summing products of the form f(x*) times the subinterval width. As the widths shrink to 0, the limit of the Riemann sum equals the definite integral (AP Calc Topics 6.2-6.3).
A Riemann sum is the official AP way to estimate the area under a curve (a definite integral) when you can't or don't want to integrate exactly. You chop the interval [a, b] into pieces (a partition), build a rectangle on each piece whose height comes from the function's value somewhere in that subinterval, and add up the rectangle areas. Each term in the sum is just height times width, f(x*) times Δx. Where you grab the height determines the type: the left endpoint gives a left Riemann sum, the right endpoint gives a right Riemann sum, and the middle gives a midpoint sum. The CED also lets the partition be nonuniform, meaning the subintervals don't all have to be the same width (LIM-5.A.2).
Here's the big payoff. A Riemann sum isn't just an estimation trick, it's the definition of the definite integral. The CED states that ∫ from a to b of f(x) dx is the limit of Riemann sums as the widths of all subintervals approach 0. So every time you write an integral, you're secretly writing 'infinitely many infinitely skinny rectangles.' That's why the dx is there. It's the Δx that shrank to nothing.
Riemann sums live in Unit 6 (Integration and Accumulation of Change), specifically Topics 6.2 and 6.3. They support learning objective 6.2.A, approximating a definite integral using geometric and numerical methods, and 6.3.A/6.3.B, interpreting and representing the limit of a Riemann sum as a definite integral. This is the conceptual bridge of the whole unit. Before Riemann sums, you only know derivatives. After them, you understand what an integral actually means, which sets up the Fundamental Theorem of Calculus and every accumulation problem that follows. The exam also expects you to judge whether an approximation is an overestimate or underestimate based on whether the function is increasing or decreasing (LIM-5.A.4), which is a classic justification point on FRQs.
Definite Integral Notation (Unit 6)
The definite integral is literally defined as the limit of a Riemann sum. Topic 6.3 asks you to translate in both directions, turning a summation into integral notation and an integral into a limit of a sum. The ∑ becomes ∫, f(x*) stays as f(x), and Δx becomes dx.
Left Riemann Sum and Right Riemann Sum (Unit 6)
These are the two endpoint flavors, and the exam loves asking which one over- or underestimates. The pattern to memorize: for an increasing function, left underestimates and right overestimates. For a decreasing function, flip it. Sketch one rectangle and you can re-derive this in five seconds.
Trapezoidal Sum (Unit 6)
A trapezoidal sum replaces each rectangle with a trapezoid, which is the same as averaging the left and right sums. It's usually more accurate, and its over/under behavior depends on concavity instead of increasing/decreasing. Concave up means overestimate, concave down means underestimate.
Partition (Unit 6)
The partition is the skeleton of every Riemann sum. On FRQ table problems the partition is handed to you as the table's x-values, and it's often nonuniform, so you can't just multiply by one fixed width. You have to compute each Δx separately.
Riemann sums show up two ways. In multiple choice, you'll identify which type of sum over- or underestimates for an increasing or decreasing function, compute a sum from a table or graph, or recognize a limit of a sum as a definite integral (and yes, one classic stem asks which method is NOT a Riemann sum; the trapezoidal sum is the answer because it doesn't use rectangles). In FRQs, the Riemann sum question is almost a yearly tradition. The 2017 (tank cross sections), 2021 (bacteria density), 2022 (melting ice sculpture), and 2023 (gas pumping rate) FRQs all gave a table of values and asked for a left or right Riemann sum or trapezoidal approximation of an integral. To earn the points you typically have to write out the sum with correct widths (watch for nonuniform partitions), state what the value means in context with units, and sometimes justify over/underestimate using the function's behavior. Calculators don't save you here; these are usually set up so you must show the hand computation.
A trapezoidal sum is technically not a Riemann sum, because Riemann sums use rectangles whose height is a single function value, while trapezoids use two values per subinterval. The CED lists them together as approximation methods, and the AP exam treats 'which of these is not a Riemann sum?' as fair game. Also, their error rules differ. Left/right Riemann sum errors depend on whether f is increasing or decreasing, but trapezoidal error depends on concavity.
A Riemann sum approximates a definite integral by adding up rectangle areas, where each term is a function value times the width of a subinterval.
The definite integral is defined as the limit of Riemann sums as the subinterval widths approach 0, which is exactly what LO 6.3.B says.
For an increasing function, a left Riemann sum underestimates and a right Riemann sum overestimates the integral; reverse both for a decreasing function.
Partitions can be nonuniform, so on table-based FRQs you must compute each subinterval width separately instead of using one fixed Δx.
The trapezoidal sum is an approximation method but not a true Riemann sum, since it uses trapezoids instead of single-height rectangles.
On FRQs, a Riemann sum answer usually needs three things: the written-out sum, the numerical value with units, and an interpretation or over/underestimate justification.
It's an approximation of a definite integral made by splitting an interval into subintervals and summing f(x*) times Δx for each piece. It's covered in Topics 6.2 and 6.3 of Unit 6, and the limit of a Riemann sum is the definition of the definite integral.
No. Riemann sums use rectangles with a single function value as the height, while a trapezoidal sum uses two values per subinterval. AP multiple choice has asked exactly this, listing left, right, midpoint, and trapezoidal and expecting you to spot trapezoidal as the odd one out.
Check whether the function is increasing or decreasing on the interval. Increasing means left sums underestimate and right sums overestimate; decreasing flips it. This justification (LIM-5.A.4) is worth points on FRQs, so state the function's behavior explicitly.
A Riemann sum is a finite approximation using n rectangles; the definite integral is the exact value you get when you take the limit as all subinterval widths shrink to 0. Per the CED, ∫ from a to b of f(x) dx equals the limit of the Riemann sum as max Δx approaches 0.
Yes, almost every year. The 2017, 2021, 2022, and 2023 exams all had FRQs giving a table of values and asking for a left or right Riemann sum or trapezoidal approximation, usually with a context interpretation and sometimes a nonuniform partition.