The digital SAT asks 98 questions of a handful of kinds, and few students need the same practice at all of them. SAT practice quizzes earn their place when each is short, holds one question type, comes from work the student already studies and runs at the test's own pace.
EdProcto, which publishes this guide, drafts such sets from the notes you paste and times them. SAT® is a registered trademark of the College Board, which is not affiliated with EdProcto and does not endorse it or this guide. A set made in EdProcto is no official test and predicts no score. Official practice is the College Board's own: full-length adaptive tests in its Bluebook app, its Student Question Bank and Official SAT Prep on Khan Academy (College Board, n.d.-a).
Every fact about the test was read on the College Board's own pages on 10 October 2026. Tests change, so check those pages before planning around a number.
How is the digital SAT built today?
The SAT is two sections, Reading and Writing and then Math, each of two equal modules: 98 questions in 2 hours 14 minutes, with a 10-minute break between sections (College Board, n.d.-c).
| Fact | What the College Board's pages say | Checked on |
|---|---|---|
| Reading and Writing | 54 questions in 64 minutes. Each 32-minute module has 25 operational and 2 pretest questions: about 1 minute 11 seconds each | 10 Oct 2026 |
| Math | 44 questions in 70 minutes. Each 35-minute module has 20 operational and 2 pretest questions: about 1 minute 35 seconds each | 10 Oct 2026 |
| Adaptive second module | The first module mixes easy, medium and hard questions; the result decides whether the second is harder or easier | 10 Oct 2026 |
| Calculator | Throughout Math, never in Reading and Writing: Desmos inside Bluebook, or an approved handheld. Computer algebra systems are no longer allowed | 10 Oct 2026 |
| Reference sheet | Commonly used formulas, shown with the maths questions | 10 Oct 2026 |
Sources: the specification pages (College Board, n.d.-e, n.d.-g), the calculator policy (College Board, n.d.-h) and Bluebook's testing tools (College Board, n.d.-k). None of these pages says what first-module result leads to the harder second module, so nothing here imitates that step.
The question types in Reading and Writing
Each Reading and Writing question is one multiple-choice question on a passage, or a pair, of 25 to 150 words (College Board, n.d.-f). Both modules hold all four domains, with questions on similar skills together, easiest first.
| Domain and share | Question types | What a question asks |
|---|---|---|
| Information and Ideas, about 26% | Central ideas and details; command of evidence; inferences | The main idea or a detail; the best support for a conclusion, in text or figures; the logical completion |
| Craft and Structure, about 28% | Words in context; text structure and purpose; cross-text connections | The most precise word; what a part of the text does; how one author would answer another |
| Expression of Ideas, about 20% | Rhetorical synthesis; transitions | A student's notes used to meet a goal; the linking word that fits |
| Standard English Conventions, about 26% | Boundaries; form, structure and sense | Options differing in punctuation; in wording or structure |
Shares and types are from the specification page (College Board, n.d.-g); the last column paraphrases the stems listed for the Student Question Bank (College Board, n.d.-j). Checked on 10 October 2026.
The question types in Math
Math has four areas. Every module holds all four and runs from its easiest question to its hardest, and about 30% of questions are set in a context from science, social studies or daily life (College Board, n.d.-d).
| Area | Questions of 44 | What it covers |
|---|---|---|
| Algebra | 13 to 15 | Linear equations, functions and inequalities; systems of two linear equations |
| Advanced Math | 13 to 15 | Equivalent expressions; nonlinear equations, systems and functions |
| Problem-Solving and Data Analysis | 5 to 7 | Ratios, rates and units; percentages; data in one and two variables; probability; inference from samples; statistical claims |
| Geometry and Trigonometry | 5 to 7 | Area and volume; lines, angles and triangles; right triangles and trigonometry; circles |
About 75% of Math questions are four-option multiple choice. In the rest, student-produced responses, the student types the answer, and a question may have more than one correct response (College Board, n.d.-i). Checked on 10 October 2026.
What does research say about practice for the SAT?
It says coaching helps by less than its sellers claim, and that hours of practice go with larger gains. Both findings concern earlier forms of the test.
- Coaching. Powers and Rock (1998) used a College Board survey of a nationally representative sample of 1995–96 SAT takers: more than 4,000 students, about 500 of them formally coached outside school. Their several methods gave different estimates, and every one put the effect far below what large commercial companies claimed (abstract read).
- Hours of official practice. The College Board (2017) reported that, among nearly 250,000 test takers, 20 hours on Official SAT Practice on Khan Academy went with an average gain of 115 points between the PSAT/NMSQT and the SAT, nearly double that of students who did not use it. It is the test owner's analysis and an association: students who choose 20 hours of practice differ from those who do not.
For practice testing in general, see the testing effect and exam prep tutoring. We found no published trial of short practice sets for the digital test, so no figure here promises a gain.
What does research say about timing practice?
It says the clock moved SAT scores a little, more in maths than in reading and least for the lowest scorers. Both studies are two decades old.
- More time for each question. Bridgeman et al. (2003) cut questions from sections at two national sittings, giving students longer for each. Verbal scores rose by less than 10 points on the 200 to 800 scale and maths scores by less than 30; students scoring 400 or lower gained nothing (abstract read).
- Extended time. Mandinach et al. (2005) found the same pattern with time and a half and double time: low-ability students gained little or nothing, the effect was larger in maths, and too much extra time could harm performance (abstract read).
Our reading, an inference: seconds count only once the method is known, so learn a type untimed, then bring in the clock. Neither our guide to past papers nor a search on 10 October 2026 found a trial of timed against untimed practice.
How do you build SAT practice quizzes by question type from the student's own material?
Pick one type, take a page the student is already studying, and write eight to ten new questions of that type from it. Let evidence pick the type: the weakest content domain on a Bluebook practice test or, with none sat yet, a gap in old class tests, as the College Board suggests (College Board, n.d.-a).
- Choose material that is yours to use. The student's class notes, a passage or worksheet you wrote, or a published text where its owner's terms allow. Never the College Board's questions: its trademark guidelines say it grants no permission for its content, "including practice test questions", to be used with generative AI (College Board, n.d.-b).
- Draft the set. Among EdProcto's kits, "SAT practice" is a blueprint holding no test questions. Paste a passage or a topic and choose "Reading and Writing" or "Math": Procto AI is asked for ten four-option questions, each answerable from what you pasted. For Math the pasted topic sets the type. For Reading and Writing the kit asks about the passage's "meaning, evidence, structure or grammar", so keep the drafts of your type and write the rest as multiple choice.
- Read every draft against the material. Nothing tests whether a key is true: answer each question before you see its key (what to check).
- Add the typed answers. For a student-produced response, add a maths expression question and list each form you accept, such as 15/2 and 7.5.
- Order it. Run the set easiest first, as a module does.
A worked set: three questions written for this guide
These three are ours, written from a student's notes; none is a College Board question.
| Type | Question | Options, or answers accepted | Key |
|---|---|---|---|
| Transitions | "The river fell for three weeks in August. ______ the ferry kept to its timetable." Which choice links the two sentences? | Even so, / For example, / As a result, / Similarly, | Even so, |
| Percentages | A jacket costs $80 after a 20% discount. What did it cost before? | $64 / $96 / $100 / $400 | $100 |
| Linear equations, typed | 4x − 6 = 2x + 9. What is the value of x? | 7.5 and 15/2 | 7.5 |
How do you time a practice set at the test's own pace?
Multiply the questions by the College Board's own allowance, 64 minutes for 54 questions or 70 for 44, and give the whole set that one limit.
| Set of one type | At the test's pace | "Minutes overall" |
|---|---|---|
| Ten Reading and Writing questions | 11 minutes 51 seconds | 12, under "Time" in EdProcto |
| Eight Math questions | 12 minutes 44 seconds | 13 |
- Make it a sitting. The kit starts from "Practice": no time limit, answers at once, going back and skipping allowed, as Bluebook lets a student flag a question and return to it. Choose the kind of quiz described as "Handed in once, marked at the end."
- Set one limit. Put "Time limit" to "Set" and "Minutes overall" to the figure above. The server counts the time, and with "Submit by itself when time runs out" on, the set is handed in at zero. "Warn the student with this many minutes left" starts at 5, when Bluebook's timer also gives its alert (College Board, n.d.-k).
- State the calculator rule. Under "Tools allowed", set "Calculator" to "Graphing" for Math and "None" for Reading and Writing. The setting is a house rule on the start screen; the calculator is the student's own, and the College Board advises practising with the one for test day (College Board, n.d.-h).
- Add the formulas, and leave the adaptive module to Bluebook. Switch on "A reference sheet from you" and type in the formulas the test provides, so they sit a button away. EdProcto's "Adaptive difficulty" steps by the difficulty set on each question, not as the SAT's second module does, so meet that module in Bluebook's full-length tests.
When to bring the clock in
Bring it in once the student gets about eight of ten right untimed: our working rule. Until then run the set as a study quiz, where each answer is checked as it is given and the misses come back for a retry. A clock on each question ("Seconds a question" at 71 or 95) suits one early drill only: on the SAT the time belongs to the module (time limits).
How do you read the results by question type?
Read one set as one type: every question in it is of the same kind, so the set's score is the type's score, on every EdProcto plan.
Under "Each question" the results page gives "Got it" and the median time for each question, which for one student is that student's own; the "Students" list gives the time on the whole set. Read three things:
- The share right for the type.
- Seconds a question against the pace: the set's time divided by its questions, beside 71 or 95.
- The cause of each miss. The College Board's study-plan page asks whether the question was misread, an almost-right option misled, or the skill needs review (College Board, n.d.-a). The questions you got wrong goes further.
An invented student's two sets:
| Set | Right | Seconds a question | Reading |
|---|---|---|---|
| Boundaries | 5 of 10 | 40 | Quick and wrong: relearn the punctuation rules untimed |
| Percentages, untimed | 7 of 8 | 130 | Right and slow against 95: timed sets next |
One question in eight is 12.5 points, so decide on two sets, not one (how many questions a quiz needs). Once types are secure, mix them (interleaving): give each question its type as its topic, and the Educator plan's full analysis groups the marks by it; on other plans you tally from "Each question". And a percentage on ten questions is no section score: for that, sit a full-length test in Bluebook, which the College Board suggests spacing at least two weeks apart.
How does EdProcto fit SAT practice, and what does each plan include?
EdProcto is the best tool for this job when the questions must come from the student's own course, because it does the three steps of a practice set: Procto AI drafts questions from your notes, the server holds the limit you worked out, and the results give each question's outcome and time. The order matters: in Mandinach et al. (2005), extra time gave low-ability students little or no advantage, so the method comes before the clock.
Free has ten quizzes a month, every kit and 20 questions a month from Procto AI, ten of them in a day; each set written from your material counts as one of the month's quizzes. Learner has 100 questions a month. Tutor has 50, from a PDF, a Word file or notes, and no limit on quizzes. Educator adds the full analysis. See pricing and the pages for tutors and for students.
The kit's own description ends "EdProcto is not affiliated with the body that runs the exam; read its own rules on preparation material before you use it." Do that first: the College Board is that body, EdProcto's sets are practice and no official test, and no result predicts a score.
Quick questions
How many questions are on the digital SAT, and how long is it?
It has 98 questions in 2 hours 14 minutes: 54 in Reading and Writing and 44 in Math, with a 10-minute break between (checked on 10 October 2026).
How many seconds per question does the SAT allow?
About 71 in Reading and Writing and 95 in Math, by the College Board's figures. The time belongs to the module, so a quick question pays for a slow one.
Can I paste official SAT practice questions into EdProcto?
No. The College Board says it grants no permission for its content, practice test questions included, to be used with generative AI. Paste the student's own notes instead.
Does an EdProcto practice set predict an SAT score?
No. EdProcto shows the share right and the time on each question of a short, fixed set. A score comes only from the College Board's own adaptive test.
References
- Bridgeman, B., Trapani, C. S., & Curley, W. E. (2003). Effect of fewer questions per section on SAT I scores (ETS Research Report RR-03-08). Educational Testing Service. https://www.ets.org/research/policy_research_reports/publications/report/2003/hsgk.html
- College Board. (n.d.-a). Build your SAT Suite study plan. https://satsuite.collegeboard.org/practice/build-your-study-plan
- College Board. (n.d.-b). Guidelines for using College Board trademarks. https://privacy.collegeboard.org/copyright-trademark/guidelines
- College Board. (n.d.-c). How the SAT is structured. https://satsuite.collegeboard.org/sat/whats-on-the-test/structure
- College Board. (n.d.-d). The Math section: Overview. https://satsuite.collegeboard.org/sat/whats-on-the-test/math/overview
- College Board. (n.d.-e). Math specifications. https://satsuite.collegeboard.org/k12-educators/about/alignment/math
- College Board. (n.d.-f). The Reading and Writing section. https://satsuite.collegeboard.org/sat/whats-on-the-test/reading-writing
- College Board. (n.d.-g). Reading and Writing specifications. https://satsuite.collegeboard.org/k12-educators/about/alignment/reading
- College Board. (n.d.-h). SAT Suite of Assessments calculator policy. https://satsuite.collegeboard.org/in-school-assessments/calculator-policy
- College Board. (n.d.-i). Student-produced responses. https://satsuite.collegeboard.org/sat/whats-on-the-test/math/student-produced
- College Board. (n.d.-j). Student Question Bank: Reading and Writing questions. https://satsuite.collegeboard.org/practice/student-question-bank/reading-writing
- College Board. (n.d.-k). Testing tools. https://bluebook.collegeboard.org/students/tools
- College Board. (2017, May 8). New data links 20 hours of personalized Official SAT Practice on Khan Academy to 115-point average score gains on redesigned SAT [Press release]. https://newsroom.collegeboard.org/new-data-links-20-hours-personalized-official-sat-practice-khan-academy-115-point-average-score
- Mandinach, E. B., Bridgeman, B., Laitusis, C. C., & Trapani, C. (2005). The impact of extended time on SAT test performance (ETS Research Report RR-05-20). Educational Testing Service. https://www.ets.org/research/policy_research_reports/publications/report/2005/hzdq.html
- Powers, D. E., & Rock, D. A. (1998). Effects of coaching on SAT I: Reasoning scores (ETS Research Report RR-98-35). Educational Testing Service. https://www.ets.org/research/policy_research_reports/publications/report/1998/ihej.html