On this page · 5 sections
AP Human Geography has no long essay question and no DBQ. If you have heard the term "LEQ," it came from APUSH, AP World History, or AP European History. The extended written task here is the free-response question, and Section II gives you three of them in 75 minutes for 50% of your exam score.
Each FRQ is one scenario broken into seven lettered parts, A through G. Every part is worth exactly one point, so each question is out of 7 and Section II is out of 21. Question 1 has no stimulus. Question 2 gives you one stimulus — a map, a graph, a data table, a photograph. Question 3 gives you two, and usually makes you work across both.
Scoring is binary and independent. A reader looks at part D, decides yes or no, and moves on without caring what you wrote in part C. There is no partial credit inside a part, no holistic writing score, no rubric row for organization. A thesis earns nothing. An introduction earns nothing. A conclusion earns nothing. The point turns entirely on whether your sentences perform the task verb the part used. Define, identify, and describe are the low rungs. Explain is the high rung, and it is the one verb that demands a stated causal link. Compare asks for similarities and/or differences between the two named things. It can be answered descriptively, and it requires a causal link only when the prompt pairs it with an explain component. Directions require complete sentences, so bullets and outlines are not scored. Label every part A through G in your response so the reader can find each answer without hunting — an unlabeled correct answer that a reader cannot locate is a lost point.
The steps
- 1
Split 75 minutes into three hard 23-minute blocks
Twenty-five minutes per question is the arithmetic, but you want a buffer, so treat 23 as the ceiling. Three 23-minute blocks run 69 minutes and leave six banked for a closing sweep. Inside a block, spend the first 90 seconds reading all seven parts and any stimulus before you write a word. That read tells you which parts overlap, so you do not spend your best example on part B and arrive at part F empty-handed. Then write for about 17 to 18 minutes and stop. If part E is not coming, leave it empty and move to F, then come back if time allows. The points are independent, so a blank E costs you exactly one point and nothing else. Students who chase a stuck part are the students who never reach Question 3, and Question 3 is a full seven points. You are allowed to answer the three questions in any order. Do the one whose units you know best first, while you are fresh and the clock is not screaming.
- 2
Circle the task verb and the number in every part before you write
Task verbs are not decoration; they are the rubric. Identify means name it, nothing more. Define means give the precise meaning. Describe means supply characteristics of the thing — what it looks like, where it is, how it is arranged — with no reason attached. Explain means state how or why, with the causal connection written out. Compare means state similarities and/or differences — read the part carefully, because it will tell you which. If it asks for one similarity, give one similarity; do not pad with a difference that earns nothing. Whatever you give must be written as an explicit comparative sentence naming both things, not two separate descriptions the reader has to compare for you. Then circle the number word. "Describe one characteristic" means one. If you give three, readers generally score the first, so a wrong first swing sinks a point your second sentence would have earned. The verb also caps you from below, not above: a correct explanation will usually satisfy a describe prompt, but a description will never satisfy an explain prompt. That asymmetry is the single most expensive thing on this exam.
- 3
Write one part per short paragraph, and open with the answer
Label the part, then make your first sentence the actual answer in the prompt's own grammar. If the part says "Describe one government policy," your first sentence names a policy and describes it. No wind-up, no restatement of the prompt, no "there are many ways in which." Two to four sentences is the right size for almost every part; anything longer is usually you talking yourself into a contradiction. That matters, because contradictory statements inside one part can cost the point even when the correct claim is sitting right there. Never leave a pronoun doing heavy lifting — "it increases" tells the reader nothing, while "the commute distance increases" is scoreable. Assume your reader is on their four-hundredth booklet and is looking for a specific, findable sentence. Give them one they cannot miss.
- 4
Force every 'explain' through the because test
Take your sentence and try to insert "because" or "which causes" into it. If nothing fits, you have written a description and you will not get the point. An explain answer has three moving pieces: a cause, a mechanism, and an outcome. "Sprawl increases driving" is a claim with no mechanism. "Sprawl separates houses from stores by distances too far to walk, and densities are too low to support frequent buses, so driving becomes the only practical option" has all three, and it earns. Use the connectors on purpose: because, so, as a result, which leads to, therefore. Readers are trained to find the causal link, and a written connector makes it findable in a second. This is also where naming a term without unpacking it fails. "This reflects the demographic transition model" is a label. Saying that falling infant mortality precedes falling fertility, so population keeps growing for a generation, is a mechanism.
- 5
Mine the stimulus out loud on Questions 2 and 3
When a part says "using the map" or "based on the data," a generically correct answer that never touches the source scores zero. Quote the source into your sentence with a number, a label, a region, or a date: "The map shows growth rates above 2.5% concentrated in Sub-Saharan Africa, while much of Europe is at or below zero." Keep your wording matched to the metric on the page. Reserve "below replacement" for total fertility rate data, meaning under about 2.1 births per woman, because a growth-rate map cannot show it. First read the title, the legend, the axis units, and the year — those four items carry most of the answerable content, and students who skip the legend routinely misread a choropleth's direction. On Question 3, expect at least one part that makes you connect both stimuli, so note in the margin what each source measures and where they disagree. When the part does not mention the stimulus, you may still use it, and it is usually the fastest source of a concrete detail. Most stimulus parts want the source detail first and the course term second; check the verb to see whether the term is actually required.
- 6
Anchor to a named place at the scale the prompt set
"In some countries" and "in many cities" are the two phrases that quietly cost the most points. Name Lagos, Detroit, Bangladesh, Portland, the Sun Belt, the Sahel. Specificity is not style here — it is the evidence the reader is checking for. Then check for scale. Some parts fix a scale explicitly — local, regional, national, or global — and when one is named, your example must sit at that scale. If a part asks about environmental effects at the regional scale, a sentence about worldwide carbon emissions is off-target even though it is true. Many parts name no scale at all, and those accept a correct example at any scale, so do not invent a constraint the prompt did not set. Keep a short mental stock of one strong example per unit — one migration case, one agricultural region, one urban model city, one boundary dispute, one industrial region — so you are never inventing a place under time pressure. Reused examples across parts are fine; recycled reasoning across parts is not.
- 7
Spend the last five minutes sweeping for the four killers
With five minutes left, stop composing and audit. Check one: is every letter A through G answered and labeled, including on Question 3? A blank part is the cheapest point on the exam to recover. Check two: does each explain part contain a written causal connector? If not, add one clause. Check three: did any part where the prompt named the stimulus actually reference the stimulus? Add the number or the region if not. Check four: does any part contain two competing answers or a sentence that contradicts your first one? Remove the weaker sentence outright, so that only the answer you want scored is left in your response — a removed contradiction can rescue a point. Do not proofread grammar. Spelling and sentence polish are not scored — but if a reader cannot read a word, that word cannot earn anything, so legibility is the one mechanical thing worth a glance. Every other second spent polishing is a second not spent on a blank part.
Worked through
Here is a Question-1 style prompt with no stimulus, worked end to end. The clock note in brackets is where you should be.
**Prompt.** Suburban sprawl has reshaped land use and transportation in metropolitan areas of the United States. A. Define urban sprawl. B. Describe one characteristic of residential land use in areas produced by sprawl. C. Describe one policy governments use to limit sprawl. D. Explain how sprawl increases automobile dependence. E. Explain one environmental consequence of sprawl at the regional scale. F. Explain how an urban growth boundary can raise housing prices inside the boundary. G. Explain one limitation of using a city's political boundaries to measure the size of its urban population.
[0:00–1:30] Read all seven. Note the ladder: A–C are low-order, D–G are all explain. Note that C and F both involve growth boundaries, so C gets the description and F gets the mechanism — do not spend the causal reasoning early.
**A.** Urban sprawl is the rapid, low-density outward spread of built-up development from a city into surrounding rural or undeveloped land. One sentence. Define is satisfied.
**B.** Residential land use in sprawl areas is dominated by detached single-family houses on large lots. Streets are curvilinear and end in cul-de-sacs, and housing is zoned separately from stores and workplaces. Characteristics only. No reason given, because none was asked for.
**C.** Some states use urban growth boundaries. Oregon requires the Portland metropolitan area to draw a line beyond which local governments will not extend sewer, water, or approval for urban development. Named place, concrete mechanism of the policy itself.
**D.** Sprawl places housing far from jobs, schools, and shops, at distances too long to walk. Low residential density also means too few riders per route mile to fund frequent bus or rail service. Because transit cannot be supported at those densities, driving becomes the only practical way to reach daily destinations, so household car ownership and vehicle miles traveled both rise. Cause, mechanism, outcome, with connectors written out.
**E.** Sprawl converts farmland and wetland at the metropolitan fringe into roofs, roads, and parking lots, which are impervious surfaces. Rain that once infiltrated the soil runs off instead, so peak flows rise in regional watersheds and flooding becomes more frequent downstream of the built edge. Regional scale, as the part demanded — a sentence about global emissions would miss here.
**F.** A growth boundary holds the supply of developable land nearly fixed while the metropolitan population keeps growing, and although boundaries are expanded periodically, the additions typically lag demand. Buyers compete for a limited number of lots, so bid prices for land rise and the housing built on that land becomes more expensive. As a result, lower-income households can be pushed to municipalities beyond the boundary. Note that this reuses the growth boundary from C but supplies entirely new reasoning, which is what the rubric wants.
**G.** Political boundaries count only residents inside the municipality, while sprawl locates millions of people in separately incorporated suburbs outside the city limits. A count based on those limits therefore excludes commuters and workers who use the city daily, so it understates the functional urban area. This is worse for older cities such as Boston, which cannot annex because incorporated suburbs surround them.
[19:00] Done, four minutes banked. Sweep: seven letters present, four causal connectors present, no contradictions.
**How this changes with a stimulus.** On Question 2, part B would likely read "Using the map, describe the pattern of population change in the metropolitan area shown." Your answer must then name what the map shows — a decline in the central city, gains at the outer ring — before you touch course vocabulary. On Question 3, expect a part that forces both sources together, such as tying a density map to a commute-time table.
This worked example sits entirely in Unit 6, but the three FRQs draw from different units across the course. The same seven-part ladder shows up over agriculture, political geography, or development. The method transfers; the content does not.
Common mistakes
Answering an explain part with a description. You name the correct process and stop, without ever writing the causal link. "Sprawl leads to more driving" states the outcome but never says why, and readers cannot award the point for an inference you left in your head.
Writing an introduction, a thesis, or a conclusion. No rubric row exists for any of them in AP Human Geography. Four sentences of set-up before part A costs you two minutes and buys zero points, and those two minutes are usually the reason part G ends up blank.
Name-dropping vocabulary instead of applying it. Writing "this illustrates the gravity model" or "this is an example of the multiplier effect" labels the concept without doing the work. Readers want the mechanism unpacked in your own sentences, with the term used correctly on the way through.
Answering a compare part with two side-by-side descriptions. If you describe subsistence farming in one sentence and commercial farming in the next, you have left the actual comparison for the reader to perform. Write the comparative sentence yourself, naming both things and the similarity or difference between them.
Hedging with a list. When a part asks for one policy and you supply three, readers generally take the first one you wrote, so a weak opening answer sinks the point even though your third option was right. Commit to one and develop it.
Ignoring the stimulus on Questions 2 and 3, or drifting off a scale the prompt named. If the prompt says "using the chart" and your answer never touches the chart's data, it scores zero however accurate it is. The same applies when a part fixes a regional scale and you answer globally.
Running out of time and leaving parts blank, almost always on Question 3. Each blank letter is a full point of the 21, and a rushed two-sentence attempt at part G is worth far more than a polished part D you spent eight minutes perfecting.
Checklist
- Can you state from memory that Section II is three questions, seven parts each, 21 points, 75 minutes, and 50% of your score?
- Can you write one describe sentence and one explain sentence about the same process, and point to exactly what makes the second one an explanation?
- Can you run any explain answer through the because test and add a causal connector when it fails?
- Can you write a compare answer as a single comparative sentence that names both things, instead of two separate descriptions?
- Can you answer a single part in under 90 seconds using two to four complete sentences that open with the answer?
- Can you name one specific real place — city, region, or country — as a ready example for each of the seven units?
- Can you tell whether a part fixes a scale, match your example to it when it does, and stop hunting for one when it does not?
- Can you read a stimulus title, legend, units, and year in under 30 seconds, then quote a number or region from it into a sentence?
- Can you finish all 21 parts with five minutes left to sweep for blanks, missing connectors, and contradictions?