Skip to content

Free-response practice

The Long FRQs

18 points2 original prompts4 scored examples

Section II of the AP Biology exam gives you six free-response questions in 90 minutes, and it counts for 50% of your score. Questions 1 and 2 are the long FRQs, worth 9 points each. Question 1 is Interpreting and Evaluating Experimental Results. Question 2 is the same task plus graphing: you draw a graph from a data table before you analyze it. The four short FRQs are worth 4 points each, so the two long questions carry 18 of the section's 34 points. Budget about 22–25 minutes per long FRQ.

On this page · 6 sections

The two questions split their 9 points differently, and the exact split varies a little from year to year. On Question 1, part (a) asks you to describe or explain a biological concept the scenario depends on (1–2 points). Part (b) tests experimental design — identify variables, controls, or a null hypothesis (1–2 points). Part (c) asks you to analyze the data: state the trend and back it with specific numbers (1–3 points). Part (d) asks you to predict what happens under a new condition and justify that prediction with a mechanism (2–4 points). A common split, and the one this guide's rubric uses, is 2/2/2/3. Question 2 always includes constructing a graph from the data table, and the graph is typically worth about 4 points. Commonly one point each goes to axis labels with units, even scaling, accurate plotting with the data series identifiable, and error bars when the data include them. In a year whose table carries no error values, that last point goes to a best-fit or trend line instead. The exact allocation varies by year; this guide's practice prompts include ± 2 SEM data, so their rubric credits error bars. Question 2's other parts — concept, data analysis, and predict-and-justify — work like Question 1's and share the remaining points. A common whole-question split is 2/4/2/1.

Readers score analytically, point by point, against a rubric. They hunt for creditable statements and ignore irrelevant extra material — but a statement that contradicts a correct answer in the same part can cancel that point, so never hedge by arguing both directions. Vague sentences never help. On Question 1, the prediction and the justification usually pay separately: a correct prediction can earn its point even if the justification fails, and a justification that just restates the prediction earns nothing. Answer every part, label your answers (a) through (d), and put a number from the data in part (c) every single time.

The rubric, in plain terms

18 points, earned line by line. Readers score what is on the page, nothing else.

Long FRQ 1, part (a): Describe and explain the underlying concept2 pts

State the relevant biological concept accurately (1 point), then explain how it works in this scenario — the mechanism, not just the name (1 point). Example: say that variation exists in the population, then explain that random mutation produced it before the selective pressure appeared.

Long FRQ 1, part (b): Experimental design2 pts

Correctly identify the requested design element, such as the dependent variable or null hypothesis (1 point), and explain the purpose of a control or a design choice (1 point).

Long FRQ 1, part (c): Analyze the data2 pts

Describe the pattern the data show, comparing treatments or groups (1 point), and support the claim with specific values from the table or graph — actual numbers, and error bars where significance is asked (1 point).

Long FRQ 1, part (d): Predict and justify3 pts

Make a directional prediction that follows from the new condition (1 point), then justify it with the biological mechanism that produces the outcome (1 point) and connect that mechanism to the evidence or to fitness, structure, or function in the scenario (1 point). Restating the prediction in new words earns nothing.

Long FRQ 2, part (a): Describe and explain the underlying concept2 pts

Same task as Question 1: state the relevant concept accurately (1 point) and explain its mechanism in this scenario (1 point).

Long FRQ 2, part (b): Construct the graph4 pts

Four separate points. Label both axes with the quantity and its units (1 point). Use even scales that fit the data (1 point). Plot every data point accurately, with the two data series identifiable by a legend or labels (1 point). Draw error bars spanning the stated interval, such as ± 2 SEM, on each mean (1 point).

Long FRQ 2, part (c): Analyze the data2 pts

Describe the pattern the data show, comparing treatments or groups (1 point), and support the claim with specific values — actual numbers, and error bars where significance is asked (1 point).

Long FRQ 2, part (d): Predict and justify1 pt

One point for a directional prediction that follows from the new condition, justified with the biological mechanism that produces it. A prediction with no mechanism, or a justification that only restates the prediction, earns nothing.

Practice prompts

Prompt 1

Selection for Antibiotic Resistance in E. coli

Researchers grew ten replicate populations of Escherichia coli in flasks of liquid growth medium. Five flasks contained a low dose of streptomycin, an antibiotic that blocks the bacterial ribosome. Five control flasks contained no antibiotic. Every 24 hours, a sample from each flask was transferred to a fresh flask of the same type. After each transfer, the researchers plated a sample from each flask onto high-streptomycin agar and recorded the percentage of cells that formed colonies (percent resistant).

Mean percent resistant cells: Day 0 — control flasks: 0.001%; streptomycin flasks: 0.001% Day 4 — control flasks: 0.001%; streptomycin flasks: 12% Day 8 — control flasks: 0.002%; streptomycin flasks: 64% Day 12 — control flasks: 0.001%; streptomycin flasks: 98%

Separately, the researchers found that in antibiotic-free medium, resistant cells divide about 15% more slowly than susceptible cells.

  1. (a)Describe the role of genetic variation in evolution by natural selection. Explain how streptomycin-resistant cells can be present in populations that have never been exposed to streptomycin. [2 pts]
  2. (b)Identify the dependent variable in this experiment. Explain why the flasks without antibiotic are a necessary control for the researchers' conclusion. [2 pts]
  3. (c)Describe the difference in how percent resistance changes over time between the streptomycin flasks and the control flasks. Support your description with specific data from the table. [2 pts]
  4. (d)On day 12, the researchers remove streptomycin from the five treated flasks and continue daily transfers in antibiotic-free medium for 12 more days. Predict how the percentage of resistant cells in these flasks will change, and justify your prediction using the growth-rate data. [3 pts]
Prompt 2

Temperature and LDH Enzyme Activity in Two Fish Species

Lactate dehydrogenase (LDH) is an enzyme of anaerobic metabolism found in nearly all animals. Researchers purified LDH from two fish species: an Antarctic icefish that lives in water near 0°C, and a desert pupfish that lives in springs near 35°C. They measured the activity of each enzyme (micromoles of product per minute) at four temperatures. Each value below is a mean of six trials, reported ± 2 SEM.

Antarctic icefish LDH: 5°C: 42 ± 4; 15°C: 55 ± 5; 25°C: 30 ± 6; 35°C: 8 ± 3 Desert pupfish LDH: 5°C: 6 ± 2; 15°C: 20 ± 4; 25°C: 48 ± 5; 35°C: 61 ± 4

  1. (a)Describe how an enzyme such as LDH increases the rate of a chemical reaction. Explain why an enzyme's activity depends on the three-dimensional shape of its active site. [2 pts]
  2. (b)Using the data in the table, construct a graph of mean LDH activity as a function of temperature for both species on the same axes. Include error bars representing ± 2 SEM for each mean. [4 pts]
  3. (c)Identify the approximate temperature at which each species' LDH is most active. Using the error bars, justify whether the two enzymes differ significantly in activity at 35°C. [2 pts]
  4. (d)Ocean warming is predicted to raise the icefish's habitat temperature over the next several centuries. Predict how the icefish population's LDH temperature optimum could change over many generations, and justify your prediction in terms of protein structure and natural selection. [1 pt]

Scored examples

Two responses per prompt — one near the top of the rubric, one in the middle — with every point accounted for. The mid response earns its score through the mistakes students actually make.

Selection for Antibiotic Resistance in E. coli · high-scoring response9/18

a) Natural selection can only act on a population that varies. Random mutations during DNA replication create genetic differences among cells, and some of those differences affect survival. A mutation that changes the ribosomal protein streptomycin targets can arise by chance in any large population, whether or not the antibiotic is present. That is why 0.001% of cells are already resistant on day 0 — the variation exists before selection, not because of it.

b) The dependent variable is the percentage of cells that form colonies on high-streptomycin agar (percent resistant). The no-antibiotic flasks are a necessary control because they show what happens to resistance frequency under identical transfer conditions without the antibiotic. Since the controls stay near 0.001%, the rise in the treated flasks must be caused by streptomycin, not by the daily transfers or by time alone.

c) In the streptomycin flasks, percent resistant climbs rapidly: from 0.001% on day 0 to 12% by day 4, 64% by day 8, and 98% by day 12. In the control flasks it stays essentially flat, between 0.001% and 0.002%, the whole time. Streptomycin kills susceptible cells, so resistant cells survive, reproduce, and take over the treated populations.

d) I predict the percentage of resistant cells will decrease over the 12 antibiotic-free days. Resistance carries a fitness cost: resistant cells divide about 15% more slowly than susceptible cells when no antibiotic is present. Once streptomycin is removed, susceptible cells have the higher fitness — they divide faster and leave more descendants in each daily transfer. Over repeated generations, susceptible cells make up a growing share of each flask, so the resistant fraction falls from 98% toward a much lower frequency.

How it scored

2/2a — Earned both points: describes variation as the raw material selection acts on, and correctly explains that resistance mutations arise randomly before exposure, citing the day 0 data.
2/2b — Identifies the dependent variable precisely and explains the control's purpose — isolating the antibiotic as the cause by holding transfer conditions constant.
2/2c — States the contrasting trends and supports both with specific values (0.001% to 98% versus flat at 0.001–0.002%).
3/3d — Correct directional prediction (decrease), justified by the 15% growth-rate cost, and connected to fitness and reproduction over generations.
Selection for Antibiotic Resistance in E. coli · mid-scoring response4/18

a) Natural selection is survival of the fittest, and populations have variation so some bacteria are stronger than others. When the bacteria were exposed to streptomycin they had to adapt, so they mutated to become resistant. The antibiotic made the bacteria develop resistance so they could survive in the flasks.

b) The dependent variable is the percent of resistant bacteria. The control flasks without antibiotic are there so the researchers have something to compare their results to.

c) The streptomycin flasks went from 0.001% resistant at the start up to 12% on day 4, then 64%, and reached 98% by day 12. The control flasks stayed at about 0.001% the entire experiment. So the antibiotic flasks became almost completely resistant while the controls did not change.

d) I predict the flasks will stay at about 98% resistant after the antibiotic is removed. Once the bacteria have the resistance mutation, it is part of their DNA and does not go away, so the population will stay resistant even without streptomycin.

How it scored

1/2a — Earned the description point for noting that populations vary. Lost the explanation point: the response says the antibiotic made the bacteria mutate, but mutations are random and exist before exposure — a documented misconception.
1/2b — Dependent variable correctly identified. 'Something to compare to' is too vague for the control point — it never says what the comparison isolates.
2/2c — Both trends described and supported with specific numbers from the table.
0/3d — Prediction is wrong: it ignores the stated 15% growth-rate cost of resistance, so no prediction point. The justification treats a fixed mutation in one cell as fixed in the population and earns nothing.
Temperature and LDH Enzyme Activity in Two Fish Species · high-scoring response9/18

a) LDH speeds up its reaction by lowering the activation energy: it binds substrate at its active site and stabilizes the transition state, so many more substrate molecules react per second. It does not change the ΔG of the reaction or add energy to it. Activity depends on the active site's three-dimensional shape because catalysis requires the substrate to fit the site's specific geometry and chemistry. If folding changes, the substrate binds poorly or not at all, and the rate drops.

b) (Graph drawn.) X-axis labeled 'Temperature (°C)' scaled evenly from 0 to 40; y-axis labeled 'LDH activity (µmol product/min)' scaled evenly from 0 to 70. Icefish means plotted at (5, 42), (15, 55), (25, 30), (35, 8) and pupfish means at (5, 6), (15, 20), (25, 48), (35, 61), each point with vertical error bars spanning ± 2 SEM. The two series are connected point-to-point and identified in a legend.

c) The icefish LDH is most active near 15°C (55 µmol/min), and the pupfish LDH is most active near 35°C (61 µmol/min). At 35°C the difference is significant: the icefish error bars span 5 to 11 and the pupfish error bars span 57 to 65. The intervals do not overlap, so the two enzymes almost certainly differ in activity at that temperature.

d) I predict that over many generations the icefish population's LDH optimum will shift toward warmer temperatures. LDH activity depends on active-site shape, and the cold-adapted enzyme loses its functional shape as temperature rises — its activity already falls from 55 at 15°C to 8 at 35°C. Individual icefish vary in LDH amino acid sequence, and some variants hold a stable active site at warmer temperatures. As the water warms, fish carrying heat-stable variants keep more metabolic function, survive, and reproduce more. Those alleles rise in frequency, so the population's average LDH optimum moves upward.

How it scored

2/2a — Earned both points: activation energy lowered without changing ΔG, and a clear mechanism linking active-site shape to substrate binding and rate.
4/4b — Earned all four graphing points: both axes labeled with quantities and units, even scales that fit the data, all eight means plotted accurately with a legend identifying the two series, and ± 2 SEM error bars on every mean.
2/2c — Both optima identified with values, and significance at 35°C justified by non-overlapping error-bar ranges (5–11 versus 57–65).
1/1d — Earned the predict-and-justify point: a directional prediction about the population's optimum (shift warmer over generations), justified through protein structure and selection — heritable variation, differential survival, and rising allele frequency.
Temperature and LDH Enzyme Activity in Two Fish Species · mid-scoring response3/18

a) Enzymes speed up reactions by giving the reaction the energy it needs to get over the energy barrier. LDH has an active site that fits its substrate like a lock and key, so if the shape of the active site changes, the substrate won't fit and the enzyme won't work.

b) (Graph drawn.) Temperature on the x-axis from 5 to 35 and activity on the y-axis from 0 to 70, with both species' points plotted correctly and connected by lines, one solid and one dashed. No units on the y-axis label and no error bars drawn.

c) The icefish enzyme works best at around 15°C and the pupfish enzyme works best at around 35°C. At 35°C the pupfish enzyme is clearly much better because 61 is a lot bigger than 8, so the difference is significant.

d) I predict the icefish LDH will not work as well as the ocean warms, because the enzyme will start to denature at higher temperatures and lose its shape. But over time the fish's enzymes will get used to the warmer water, so each fish will adjust its LDH to work at the new temperature during its life.

How it scored

1/2a — The shape point is earned through the active-site explanation. The rate point is lost: enzymes do not give energy to a reaction, they lower the activation energy barrier — a documented misconception.
1/4b — Earned only the scaling point (even scales that fit the data). Lost the axis-label point because the y-axis has no units. Lost the plotting point: solid versus dashed lines make the two series distinguishable from each other, but with no legend or labels the reader cannot tell which line is the icefish and which is the pupfish — the rubric requires the series to be identifiable, not just distinct. Lost the error-bar point because the required ± 2 SEM bars are missing.
1/2c — Both optima correctly identified with temperatures. The significance point is lost: comparing 61 to 8 without using the error bars does not justify statistical significance.
0/1d — Not earned. The question asks for a prediction about the population's temperature optimum over many generations; the response predicts only that the current enzyme will perform worse. Its mechanism is also wrong for the point: individual fish 'getting used to' warm water is acclimation within a lifetime, not natural selection acting on heritable variation.

Where points die

Writing that bacteria, insects, or enzymes mutate in response to a stressor — 'the antibiotic made them resistant.' Mutations are random and happen before the selective pressure appears. The stressor does not create the variant; it kills the alternatives. Selection changes allele frequencies in the population — it never changes an individual. Readers deny the explanation point instantly for directed mutation.

Saying an enzyme adds energy to a reaction, raises the energy of the reactants, or changes the reaction's ΔG. Enzymes only lower the activation energy by stabilizing the transition state. The free energy of reactants and products — and therefore ΔG and the equilibrium — is unchanged. The reaction just reaches the same endpoint faster.

Naming the wrong type of selection when a trait distribution shifts — calling a one-direction shift 'stabilizing' or 'balancing' selection. Match the pattern to the fitness winners. Directional: one extreme is favored and the mean shifts. Stabilizing: intermediates are favored and variation shrinks. Disruptive: both extremes are favored and intermediates decline. Balancing selection maintains multiple alleles, as in heterozygote advantage.

Crediting natural selection for every frequency change — including one the stem says happened 'by chance' in a small population. Genetic drift, including the founder effect and bottlenecks, changes allele frequencies randomly, and its effects are strongest in small populations. Invoke selection only when the trait demonstrably affects survival or reproduction. If the stem says 'by chance' or gives a tiny founding population, write drift.

Claiming two treatments differ significantly just because their means differ, without touching the error bars. Part (c) significance points require the error bars. State the interval each bar spans and say whether the intervals overlap. Non-overlapping ± 2 SEM bars support a real difference; overlapping bars mean you cannot conclude one.

Justifying a prediction by restating it — 'resistance will drop because there will be fewer resistant cells.' On Question 1, the prediction and the justification are usually separate rubric points; on Question 2 they often share a single point. Either way, the justification must name the mechanism: which cells have higher fitness and why, what the enzyme's structure does at that temperature, what the gradient drives. If your 'because' clause could be deleted without losing information, it earns nothing.

Score your own response

0 / 18

Be the harshest reader in the room: award a point only when your response does what the line says, on the page — not what you meant to write.