Duolingo English Test: The Complete 2026 Preparation Guide
Sixty minutes. That's the entire window between opening your laptop and having a number that could decide whether a university even looks at your application. No proctor in the room, no test center, just you, a webcam, and roughly 45 to 55 tasks fired at you by an algorithm that's constantly recalibrating how hard the next question should be. It's a strange exam to prepare for, because half the battle isn't English at all — it's understanding a system that's evaluating you while you're still figuring out what it wants.
I've dug through the official Duolingo materials and a stack of 2026 prep guides to pull together what actually moves the needle, not the generic "practice every day!" advice that could apply to literally any test on earth.
What Actually Happens During the Test
The Duolingo English Test (DET) runs in three chunks, and knowing the shape of it in advance removes a surprising amount of anxiety.
- Onboarding (5 minutes) — ID verification, webcam and mic checks, environment scan. Unscored, but mess it up and your results won't get certified.
- Adaptive Test (45 minutes) — the scored core. Questions pull from listening, reading, writing, and speaking, and difficulty shifts in real time based on how you're doing.
- Writing & Speaking Sample (10 minutes) — doesn't touch your numeric score, but universities can view these samples directly, so a lazy 40-word answer here can undercut a strong overall number.
The adaptive section pulls from a bank of roughly 19 distinct task types, grouped into a few families: word-recognition tasks like Read and Select or Listen and Select, cloze-style tasks like Read and Complete, and production tasks like Read Aloud, Speak About a Topic, and Write About a Topic. You won't know which specific tasks you'll get — that's the point of adaptive testing — but you can absolutely drill every category until none of them feel unfamiliar.
Worth understanding the mechanics here, because it changes how you should react mid-test. Adaptive algorithms like this one adjust difficulty question-by-question based on your running accuracy, which means a shaky start doesn't necessarily doom the rest of the section — the system recalibrates rather than locking you into a low band permanently. What it does punish is inconsistency: getting an easy word-recognition item wrong because you rushed costs you more, proportionally, than struggling honestly with a genuinely hard one. So the practical instinct to build is patience on the easy-looking tasks, not just speed. A missed "gimme" early on can pull the algorithm toward feeding you a lower difficulty ceiling for the rest of the section, which caps how high your eventual subscore can climb even if you nail everything after it.
Here's the part people underestimate: the test got a format refresh in August 2025 that changed how certain tasks feed into subscores, even though the overall 60-minute, three-section shell stayed put. If your prep material is older than that, some of its task-weighting advice is stale.
How the Scoring Actually Works
Duolingo scores land somewhere between 10 and 160, moving in steps of 5, and the overall number is just the average of four subscores rounded to the nearest 5: Literacy, Comprehension, Conversation, and Production. Each one carries equal weight — 25% — so you can't coast on strong reading and ignore speaking.
The average overall score sits around 110. The 25th percentile lands near 95, and the 75th percentile is around 125 — which tells you the real competitive band is narrower than most people assume.
A "good" score depends entirely on what you're applying for, and this is where a lot of test-takers waste energy chasing a number nobody asked them for.
| DET Score | Rough IELTS Equivalent | Rough TOEFL iBT Equivalent | Typical Use Case |
|---|---|---|---|
| 100–110 | 5.5–6.0 | ~70–75 | Foundation/pathway programs |
| 115 | 6.0 | ~80 | Common undergrad minimum |
| 120 | 6.5 | ~90 | Standard graduate requirement |
| 130–135 | 7.0 | ~100+ | Competitive/selective programs |
| 140+ | 7.5+ | ~105+ | Highly selective, some scholarships |
The practical takeaway: don't train toward "as high as possible." Find your target school's stated minimum, add 5–10 points as a buffer, and stop there. Overshooting by 20 points burns weeks you could spend on your actual application.
There's a subtler trap worth flagging, too: some programs publish a minimum overall score but also enforce hidden subscore floors — a program might accept an overall 115 but quietly reject anyone whose Production subscore falls below 100, even if Literacy and Comprehension are carrying the average. If a program's admissions page only lists one number, email the international admissions office and ask whether subscore minimums apply before you assume a single average is the whole story. It's a five-minute email that can save you a wasted retake.
Building a Study Timeline That Fits Your Level
Prep length depends on where you're starting, not how much time you feel like you should spend. Three realistic tiers, based on patterns across current DET guides:
- Beginner (CEFR A2–B1, DET under 95): plan on 1–2 months of near-daily practice, prioritizing vocabulary breadth and basic sentence construction before touching timed drills.
- Intermediate (DET 95–115): 4–6 weeks is usually enough if you're combining timed practice with feedback, not just passive review.
- Experienced / near-target (DET 115+, retaking to nudge a score up): under 4 weeks, focused entirely on your single weakest subscore rather than a full review.
That last point matters more than it sounds. If you've already taken the official test once, you have subscores in hand — use them. Start with your lowest subscore, not your favorite one. Most test-takers instinctively over-practice reading (it feels comfortable) while their Production subscore, which blends writing and speaking, quietly drags the average down.
A mistake I see repeated across forums: treating Duolingo's free official practice test as sufficient prep. It gives you a single estimated score with no itemized feedback on why you lost points — useful as a baseline, useless as a study tool. Pair it with a scored practice platform or a native-speaking friend who'll actually critique your Read Aloud pacing and Write About a Topic structure.
The Vocabulary-and-Exposure Layer
Underneath the task-specific drilling, there's a slower-burn skill: raw vocabulary depth, since several tasks (Read and Select, Read and Complete) are essentially testing whether a word "looks real" to you. Reading varied material — not just textbooks, but magazine features, opinion columns, even Reddit threads in English — builds the pattern recognition those tasks reward. Aim for genuinely new words, not ones you've half-learned three times already.
A Practical Playbook by Question Type
Different task families need different practice, and lumping them together under "study English" wastes time.
For word-recognition tasks (Read/Listen and Select): speed matters more than depth here. These are pass/fail on individual words, so practice scanning rather than deep-reading each option.
For Read Aloud and Speak About a Topic: the guidance that keeps showing up across prep resources is simple but ignored constantly — if you're given a 30–90 second window, use close to the full 90, not the minimum 30. A thin, minimal answer signals lower fluency even if every word is correct. Record yourself, play it back, and listen for filler words and unnatural pausing before you judge grammar.
To make that concrete: a weak Speak About a Topic answer on "describe a place you'd like to visit" sounds like "I want to visit Japan. It is beautiful. I like the food." — three flat sentences, maybe 15 seconds, no elaboration. A stronger answer at the same skill level extends the same idea with reasons and specifics: "I'd like to visit Kyoto because it mixes traditional temples with modern city life, which I find really interesting. My cousin visited last year and said the food markets alone were worth the trip, so I'd want to spend at least a few days just exploring those before seeing the historic sites." Same vocabulary difficulty, but it fills the time window, shows connected reasoning, and gives the Conversation subscore something to actually measure.
For Write About a Topic and Interactive Writing: these expect roughly 50–100 words with actual structure — a claim, a reason, an example — not a stream of loosely connected sentences. Practicing five of these a week, with someone else reading them cold, exposes structural gaps faster than grammar drills do.
For Listen and Summarize / Dictation: this is where typing speed becomes a real bottleneck. If you're a slow typist, that's a solvable problem separate from your English ability, and it's worth 20 minutes a day on a typing trainer before test week.
One counter-intuitive point: don't over-optimize for "sounding academic." Duolingo's rubric rewards clarity and coherence over vocabulary showing off — a well-structured, simple sentence beats a tangled attempt at a fancy one that falls apart halfway through.
Test-Day Logistics That Can Sink an Otherwise Good Score
This is the part nobody studies for, and it's the part that actually disqualifies people. The DET is proctored by AI and recorded on video for human review, and violations here can get your result flagged as "not certified" — meaning universities never see it, no matter how well you performed.
- Close every other application, including browser tabs and background grammar checkers (Grammarly included).
- No headphones — built-in speakers and mic only.
- Keep your full room visible: no other people, no notes, no second monitor in frame.
- Stable internet is non-negotiable; a mid-test disconnect can end your session.
- Have a valid government-issued ID ready before you start.
A few edge cases worth planning around rather than discovering mid-test: dim or backlit rooms (sitting with a window behind you) can make face-visibility checks fail, so face a light source instead of your back to it. Shared apartments and thin walls are a common problem too — background noise from a roommate's call doesn't automatically flag a violation, but it can degrade your speech-recognition accuracy on Read Aloud tasks, so book a genuinely quiet block rather than assuming "quiet enough" will do. And if your internet is inconsistent, test it with a video call the same time of day you plan to test — household Wi-Fi often dips during evening peak hours in ways a quick daytime check won't reveal.
Results usually arrive within 48 hours, or 12 hours if you pay roughly $40 extra for Faster Results. As of 2026, a single test runs $70, with a two-test bundle at $118 (works out to $59 each) — genuinely cheaper than IELTS or TOEFL, especially once you factor in that score reports go to unlimited universities for free. Scores stay valid for two years, and Duolingo caps how many times you can purchase a retake in a given window, so treat each attempt as one you actually prepared for, not a free dry run.
Bottom Line
- Get your subscores first if you've already tested once, and spend 70% of remaining prep time on your weakest one — usually Production.
- Time-box your speaking answers to the full window, not the minimum, since thin answers read as lower fluency regardless of grammar accuracy.
- Match your target score to your actual program's stated minimum plus a small buffer, not to some abstract "as high as possible" ceiling — and check for hidden subscore floors first.
- Rehearse the technical setup — camera framing, closed apps, stable internet — as seriously as you rehearse vocabulary, since violations can void an otherwise strong score.
- Use scored, feedback-driven practice tools instead of relying on the free official practice test as your only prep material.
Frequently Asked Questions
How many times can I retake the Duolingo English Test?
Duolingo limits how many times you can purchase the test within a rolling period, and most guides cite a cap tied to a monthly purchase limit. Each retake costs full price unless you bought a bundle in advance, so retaking casually without targeted prep is an expensive habit.
Is the Duolingo English Test easier than IELTS or TOEFL?
Not easier, just faster and less format-heavy — one hour versus two to three, and no essay-planning stage or separate listening module. The content difficulty tracks the same CEFR framework as IELTS, so a DET 120 and an IELTS 6.5 reflect roughly the same actual English ability.
Does my Writing and Speaking Sample affect my score?
No, the samples in the final 10-minute section don't factor into your 10–160 numeric score. Universities can still view them directly, though, so submitting a rushed or minimal response can undercut a strong adaptive-section score in a reviewer's eyes.
What's a good Duolingo score for graduate school?
Around 120 is the most common graduate-program requirement, roughly equivalent to an IELTS 6.5. Competitive or research-heavy programs sometimes ask for 130+, so always check the specific program page rather than assuming a universal cutoff.
Can I use notes or a dictionary during the test?
No. The AI proctoring and video review flag visible notes, paper, or a second screen, and violations can get your entire result marked uncertified. Treat the test environment exactly like an in-person exam room, even though you're taking it from home.
How is the Duolingo English Test scored if it's adaptive?
The algorithm adjusts question difficulty in real time based on your accuracy, then converts your performance across all four skill areas into subscores for Literacy, Comprehension, Conversation, and Production. Your overall score is the average of those four, rounded to the nearest 5 on the 10–160 scale.
