AP Biology Unit 7 Practice Test: Natural Selection MCQs (Free PDF)

Here's a free AP Biology Unit 7 practice test: 20 AP-style MCQs on natural selection, seven data-based stimulus sets across topics 7.1 through 7.12, and a full rationale for every answer, all aligned to the 2025 CED.

Below, three mistakes I see students make on Unit 7 and how to avoid them, plus one worked example so you can see the reasoning.

Unit 7 is the heaviest unit on the paper at 13-20% of the exam (see the full 8-unit breakdown for weightings on the rest), spanning Hardy-Weinberg population genetics through cladistics, evidence for evolution, speciation and the origin of life.

AP Biology's Section I is stimulus-heavy, and Unit 7 stimuli lean hard on quantitative and comparative data. Expect:

• Recapture-rate tables that test directional selection (rock pocket mice, industrial melanism-style examples that don't rely on the retired peppered moth)
• Hardy-Weinberg tables where you're given raw genotype counts and asked for an allele or genotype frequency
• Antibiotic-resistance-over-time tables
• Molecular vs morphological data tables (cytochrome c difference matrices, derived-character cladogram tables)
• Speciation scenarios (guppies in different streams, finches with changing beak depth)

The practice pack mirrors that mix: 13 of the 20 questions sit on a data-table stimulus. There's a rock pocket mice recapture table testing directional selection on dark lava, a beetle genotype-count table testing whether you can pull q out of raw counts, an antibiotic-resistance-over-time table, a cytochrome c amino-acid difference matrix, a derived-character-state cladogram table, and a finch beak-depth-over-20-years table.

If you know the stimulus formats, most of Unit 7 becomes reading practice as much as biology.

Every year I see students walk into Unit 7 MCQs making the same three moves. All three show up in recent Chief Reader Reports, so spot them before the exam does.

The 2025 Chief Reader Report Q4 flagged this as one of the top Unit 7 misconceptions: "Changes in the environment cause new traits to develop." Students describe evolution as goal-directed: polar bears "developed" white fur because they "needed" camouflage, giraffes "grew" longer necks to reach leaves. This is Lamarckian and it's always wrong on AP Biology.

Natural selection acts on existing genetic variation that arose randomly through mutation. A mutation for slightly lighter fur happened by chance; polar bears carrying that variant survived and reproduced better in a snowy environment, so the allele's frequency rose over generations. The environment selects for pre-existing variation; it doesn't cause new traits to appear.

On MCQs, any answer that uses the phrasing "developed X to help them Y" or "needed X" is almost always the trap answer.

The classic Hardy-Weinberg trap. The homozygous recessive frequency in the population is q², not q. Students see "16% of the population expresses the recessive phenotype" and plug 0.16 in as q, when they should take the square root first (q = 0.4). The correct downstream values are p = 1 − q = 0.6, heterozygote frequency = 2pq = 0.48 and homozygous dominant = p² = 0.36. Skipping the square-root step gives p = 0.84 and p² ≈ 0.71, which are wrong. The catch: p² + 2pq + q² still sums to 1 with the bad numbers, because the arithmetic is internally consistent. The check that catches the mistake is comparing q² against the recessive phenotype frequency the stem gave you: if q² doesn't match 0.16 here, you plugged in wrong.

The reliable workflow:

1. Frequency of homozygous recessive = q².
2. Take the square root to get q.
3. Then p = 1 − q.
4. Then compute the other genotype frequencies.

Verify by checking that q² matches the recessive phenotype frequency the stem gave you.

Two species that appear next to each other on a phylogenetic tree are not necessarily most closely related; that depends on where their branches share a common node, not where the tip labels sit on the page.

The 2023 Chief Reader Report Q5 flagged this along with the related misconception that a branch on a cladogram represents an organism "evolving into" the next one. Nodes actually represent common ancestors from which two sister groups diverged.

On MCQs, the reliable move: pick two species, trace each branch backward until they meet at a node; that node is their most recent common ancestor. The species whose common ancestor is more recent are more closely related, regardless of how the tips are drawn.

Here's a question straight from the Unit 7 pack. The stimulus: A population of 500 beetles was sampled for a single gene locus with two alleles, A and a, where A is dominant. Genotype counts were recorded.

The stem: 'Using Table 2, what is the frequency of allele a (q) in this beetle population?'

• (A) 0.09
• (B) 0.30
• (C) 0.50
• (D) 0.70

Answer: B. Count the a alleles: each aa contributes 2 (45 × 2 = 90), each Aa contributes 1 (210 × 1 = 210). Total a = 300 out of 2N = 1,000 total alleles, giving q = 0.30. A is the fraction of aa homozygotes (45/500 = 0.09). C is the value expected under allelic equality. D is p, the frequency of A.

When the stem tells you what fraction of the population has the recessive phenotype, that fraction is q². Take the square root first, then compute p, then the genotype frequencies. When the stem says "5% of the population is a carrier" or gives a heterozygote frequency directly, that's 2pq, which is harder to work with; you'll usually need a second piece of information to solve.

When the stem asks whether a population IS in Hardy-Weinberg equilibrium, check the five assumptions like a checklist:

• No mutation
• Random mating
• No gene flow
• Large population size
• No selection

If any of the four besides random mating is violated, allele frequencies will change. Non-random mating (assortative mating, inbreeding) is the one that shifts genotype frequencies (more homozygotes, fewer heterozygotes) without changing allele frequencies by itself, so a population can be out of Hardy-Weinberg equilibrium and still have stable p and q. The equation itself is on the AP formula sheet too, so no need to memorise it cold.

MCQs regularly ask "which of the following is evidence FOR evolution" versus "which is a MECHANISM of evolution". These are different categories.

• Evidence is observations that show evolution has occurred: fossils, homologous structures, embryology, biogeography, molecular data like DNA and protein sequence comparisons.
• Mechanisms are the processes that shift a population's allele or genotype frequencies. Four of them change allele frequencies directly: natural selection, genetic drift, gene flow and mutation. Non-random mating is the fifth thing that violates Hardy-Weinberg equilibrium, but it changes genotype frequencies (more homozygotes, fewer heterozygotes) without changing allele frequencies on its own.

The 2025 Chief Reader Report Q4 saw students conflate "changes in DNA sequences" (evidence) with "evolution" (the process itself). Sort each answer choice into evidence-or-mechanism before you pick.

Speciation MCQs almost always ask you to name a reproductive isolating mechanism.

Prezygotic barriers prevent mating or fertilisation from happening at all:

• Habitat
• Temporal (different mating seasons)
• Behavioural (different mating dances)
• Mechanical (incompatible anatomy)
• Gametic (sperm can't fertilise egg)

Postzygotic barriers act after fertilisation:

• Reduced hybrid viability (embryo dies)
• Reduced hybrid fertility (mule sterility)
• Hybrid breakdown (F2 generation has problems)

On MCQs the trap is usually the answer that names the wrong side of fertilisation (a prezygotic answer for a postzygotic mechanism, or the other way around), so sort each option into pre- or postzygotic first, then match to the stem.

A 3-pass method works well for a unit-length pack like this one.

• First pass, untimed: give yourself about 90 seconds per question, work through all 20, and use the rationales to understand every question you got wrong or guessed on. This is a learning pass, not a scoring pass.
• Second pass, timed: 30 minutes for 20 questions matches the AP pace of 1.5 minutes per MCQ. Circle any answer you're not confident on and move on. This is where you build exam pacing.
• Third pass, a week later: retry the questions you missed on the first two passes without looking at the answers, and see whether the reasoning has stuck.

Once the MCQ pace feels comfortable, our mock exam guide walks through pacing a full three-hour paper.