SAT Intel Report · May 2026
May 2026 SAT: easy-looking math, punishing reading
Five things this test kept doing, each with a worked example and what to practice. Math felt fine and Reading didn’t, so most of what you lost was probably in the second half.
Was the May 2026 SAT hard?
It depends which half you mean, and that is why people disagreed about it. Math was comfortable, with a couple of function questions at the end of Module 2 that weren’t. Reading and Writing is where it hurt, especially the evidence questions. If you found this one easy, you probably found it easy in math.
Weaken questions: attack the link, not the convenience
The hardest Reading question asked you to weaken a proposal. Every proposal has two parts, a practical one and a logical one, and you’re only allowed to attack the second. People went for the wrong part.
What one looked like
A compound is easier to collect from clams than from salmon. Researchers propose clams as a proxy for salmon levels, and you pick the finding that weakens it.
Where people went wrong
The finding about salmon needing special gear and permits. It looks relevant because the whole proposal started from convenience. But that finding makes the proxy more attractive, not less.
What it actually is
The finding that clam levels don’t track salmon levels. A proxy only works if measuring one predicts the other, so breaking that link kills the proposal. Convenience was never the part doing the work.
How to prep it
Split every proposal into why it is convenient and why it should work. Then remember you can only attack the second. Cross out any option about cost, access or effort before you read the rest.
Rewritten from the May 2026 SAT: same idea, new words and numbers
Because monitoring contaminant loads in mobile, fast-swimming fish is costly and labor-intensive, ecologists have long searched for stationary organisms whose tissues might stand in for them. Studying a river system in which both freshwater crayfish and brown trout accumulate the industrial flame retardant HBCD, Priya Anand and colleagues observed that crayfish, which can be trapped by the hundreds in a single afternoon, yield tissue samples far more cheaply and quickly than trout do. On that basis the team proposed that biologists could gauge HBCD contamination in the trout population simply by measuring the concentrations present in crayfish collected from the same stretch of river. Which finding, if true, would most weaken the proposal that crayfish can serve as a proxy for HBCD levels in trout?
Support questions: the finding can be an absence
Two questions asked which finding supports a hypothesis about timing. The answer was that something is missing from the record. You look for a finding that shows a thing, so it’s easy to skip the option saying nothing was found.
What one looked like
The hypothesis is that a group of languages spread with the arrival of farming. You pick the finding that supports the dating.
Where people went wrong
The finding that people in the region eat mostly cultivated crops today. It is about the same crop and the same place, so it reads as relevant. It says nothing about when farming started.
What it actually is
Excavations across the region found no trace of cultivation older than about 3,500 years. An absence is real evidence about timing, because it puts a floor under the date. The hypothesis needed a date, and that is what this gives it.
How to prep it
Write down what kind of fact the hypothesis needs before you read the options. A date, a cause, a comparison. If you decide it needs a date, keep the options about absence in play, because that’s usually how a date gets fixed.
Rewritten from the May 2026 SAT: same idea, new words and numbers
The languages of the Katukinan family are spoken across a broad expanse of the western Amazon basin, and the linguist Mariana Vasquez has argued that their present-day distribution reflects a single, comparatively recent dispersal of farming communities rather than a gradual diffusion spread over many millennia. Because reliable manioc cultivation would have permitted rapid population growth and outward expansion, Vasquez contends that the languages spread only after farming had taken firm hold in the region. If her account is correct, then the archaeological record ought to show that intensive manioc agriculture in the area is no older than the dispersal she proposes. Which finding, if true, would most directly support Vasquez's hypothesis about the languages' dispersal?
Parabolas: two equal outputs fix the axis of symmetry
The end of Module 2 had two must-be-true questions about functions. Both gave you statements to test and offered Neither as an option. People guessed instead of working out the value of a.
What one looked like
g(x) = ax squared plus 10x plus c, and g(-7) = g(1). Statement I says a is greater than 2. Must it be true?
Where people went wrong
Saying yes, because the parabola opens upward so a is positive, and 2 is a small number. Positive doesn’t mean bigger than 2. Nothing has been worked out yet.
What it actually is
Equal outputs at -7 and 1 put the axis of symmetry halfway between, at -3. The axis is at minus 10 over 2a, so that equals -3 and a is 5/3. Since 5/3 is below 2, statement I is false.
- axis of symmetry two equal outputs sit either side of it, so the axis is halfway between them. Outputs equal at -7 and 1 put it at -3.
- axis from the equation minus b over 2a. With b of 10 and the axis at -3, a comes out at 5/3.
- percent increase add the original 100 percent first. A rise of 2,900 percent is a factor of 30, not 29.
- percent less than the base is whatever comes after than. Dividing by the wrong one of two numbers is the usual slip.
- Neither as an option when a must-be-true question offers it, work the constant out rather than checking the statements loosely.
How to prep it
Turn every condition into a number before you judge a statement. Equal outputs mean an axis, an axis means a value for a. When Neither is on offer, you can be fairly sure at least one statement is designed to look forced.
Rewritten from the May 2026 SAT: same idea, new words and numbers
The functions f and g are defined by the given equations, where x ≥ 0. For which of the defined functions does the minimum value of the function, where x ≥ 0, appear as a constant or a coefficient in the equation that defines it?
I. f(x)=15(0.2)x+4
II. g(x)=5(3)x+4
Percent less than: the base comes after the word than
One Module 2 question stacked two huge percent increases and then compared the results. The arithmetic’s easy. Choosing which number goes on the bottom is the whole question.
What one looked like
A habitat starts with equal numbers of prey and predators. Over a week the prey rise by 2,900 percent and the predators by 350 percent. The predators are what percent less than the prey?
Where people went wrong
Answering 15. Start both at 1 and the prey reach 30 and the predators 4.5, and 4.5 over 30 is 0.15. That’s the fraction they are, not the amount less, so the answer is one minus it.
What it actually is
85 percent. The prey go to 30 and the predators to 4.5, so the gap is 25.5. Divide by the prey, since prey follows the word than, and 25.5 over 30 is 0.85.
How to prep it
Underline the words after than and make that your denominator. Then decide whether the question wants the fraction or the gap. Those two decisions are worth more than any amount of speed on the arithmetic.
Vocab: the blank described how it happened
- incremental
- volatile
- opaque
- expunge
- beneficiary
Vocab was the gentler part of Reading this time. The one that caught people needed a word for the manner of a change rather than for the change itself, and every trap described some other property.
What one looked like
A passage says a change happened at different places and at different times rather than all at once. The finalists were incremental, volatile and opaque.
Where people went wrong
Volatile. It’s a word about change, and the sentence is about change, so it feels close. Volatile means unstable and unpredictable, and nothing here is unpredictable.
What it actually is
Incremental, meaning in steps. Different places at different times is the definition of happening in increments. Opaque is about being hard to see through, which the sentence never raises.
How to prep it
Work out which property the blank is about before you read the options: size, speed, manner, clarity or value. Then keep only the words describing that property. A word can be about your topic and still be about the wrong property.
What to work on before your next test
- Command of Evidence, split ten proposals into their practical half and their logical half before reading any option. Back to the example
- Command of Evidence (Data), name the kind of fact the hypothesis needs, on twenty evidence questions, before looking at the choices. Back to the example
- Percentages, underline the words after than and set the denominator from them, twenty times. Back to the example
- Nonlinear Graphs, turn conditions into values for the constants rather than judging statements by eye. Back to the example
- Words in Context, label the property each blank is about before predicting a word. Back to the example
Decide what kind of thing the answer has to be
Every theme here came down to one decision you make before the work. Whether the proposal’s logic or its convenience is up for attack. Whether the hypothesis needs a date or a cause. Which number goes on the bottom. Which property the blank describes.
So make that decision out loud before you start, then answer. It takes you a couple of seconds and it’s the difference between a hard question and a slow one.
Practise the 20 questions modeled on the May 2026 SAT
Original questions written around what the May 2026 exam tested, tagged by skill and difficulty, each with a worked solution.
Students split almost evenly on this test. The split tracked which half they meant, so both halves get themes here.The reading of it is ours. We hold no scoring data, we can’t identify unscored questions, and every practice question here is written from scratch.
Questions about what was on this test
What math was on the May 2026 SAT?
Comfortable for the most part. The two that people singled out were both must-be-true function questions at the end of Module 2, where the honest answer was Neither. Alongside those came stacked percent increases, a two-way table, a linear function and a triangle question.
What vocab words were on the May 2026 SAT?
People reported incremental, volatile, opaque, expunge and beneficiary. None of them are obscure. The hard one described the manner of a change. So a word about instability or clarity was the wrong kind of word for you to pick, rather than the wrong strength.
How should I practice the May 2026 question types?
Spend your time on the evidence questions, since that is where this form did its damage. For each one, write down what kind of fact would settle the claim before you read the options. Do that twenty times and the weaken and support questions stop looking alike.