Learning Methods · 9 min read · 2026-08-31
Does Duolingo Actually Work? What the Research Says
"Does Duolingo actually work?" is one of the most-searched questions in language learning, and the honest answer has been buried under two loud extremes: marketing that implies you'll be chatting away in weeks, and cynics who insist the little owl teaches you nothing. The good news is that we don't have to guess. Duolingo has been studied in proper, peer-reviewed research — including head-to-head comparisons with real classrooms — so we can replace opinion with evidence.
The short version: the research shows Duolingo genuinely works for the foundations of a language — beginners using it reach reading, listening, vocabulary, and grammar levels comparable to classroom students — but the same research does not show it producing confident, spontaneous speaking or advanced fluency on its own. In other words, it's an evidence-backed way to build the base, not a complete path to the finish line. Here's what the studies actually found, and how a beginner should use that.
What the research actually found
The strongest evidence comes from controlled studies that gave learners real tests before and after, rather than relying on how people felt they were doing. In a study published in Cambridge's Studies in Second Language Acquisition, beginner-level French learners were split into three groups — classroom only, Duolingo only, and classroom plus Duolingo. All three groups showed significant improvement across nearly all measures, and the size of their gains was similar. Read that again: the Duolingo-only group improved by roughly the same amount as the group sitting in a classroom.
That isn't a one-off. Duolingo's own efficacy research hub and independent work point the same way for beginners: earlier studies found Duolingo instruction was about as effective as classroom instruction for grammar and vocabulary after modest weekly use, and that beginner Spanish and French learners reached reading and listening skills comparable to university students studying those languages over several semesters. The consistent thread is that for the foundational, receptive skills — understanding words and simple sentences you read and hear — Duolingo performs on par with formal instruction. For a free app you use in spare minutes, that's a genuinely strong result.
Where Duolingo genuinely delivers
Line up the research and the app's strengths cluster in one clear place: the beginning. Duolingo is effective at introducing your first few hundred to few thousand words, drilling basic grammar patterns until they feel familiar, and building your ability to recognise and understand simple written and spoken language. These are exactly the skills the studies measured and found comparable to a classroom. If your goal right now is to stop staring at a language you can't read at all and start recognising words and structures, the evidence says Duolingo will get you there.
There's a second strength the studies don't measure directly but every learner feels: the habit. Duolingo is engineered to get you to show up daily, and daily contact with a language is one of the biggest predictors of progress. An app that keeps a beginner coming back for the first three months is doing something valuable even before you count the vocabulary. This is why we still rate it as a reasonable on-ramp in is Duolingo still good for beginners in 2026 — the early value is real, and now it's backed by data.
Where the evidence runs out — speaking and fluency
Here's the part the marketing skips: the research that shows Duolingo working also quietly maps its ceiling. The gains cluster in reading, listening, vocabulary, and grammar — the skills of understanding language. What the studies don't show is Duolingo producing confident, spontaneous speaking, or carrying learners to advanced fluency. Tapping word tiles and repeating set phrases builds recognition; it doesn't build the ability to compose your own thoughts in real time or react to a real person who says something you didn't expect.
This is why the classic complaint — "I did a 500-day streak and still can't hold a conversation" — is both true and misread. It's not proof the app does nothing; it's proof the app is good at one half of the job and weak at the other. The missing half is real output: speaking, and handling unscripted, real-world language. We unpack this ceiling in more depth in why language apps stall before fluency. The takeaway isn't "quit Duolingo" — it's "know what it can't do, so you can cover that gap on purpose."
Why "Duolingo vs. a classroom" is the wrong fight
Notice which group did best in that French study: often it was classroom plus Duolingo. That's the real headline. The research isn't telling you to pick a side between the app and formal learning — it's showing that they cover different things and stack well together. Duolingo delivers cheap, daily, flexible input and vocabulary; a class or tutor delivers live speaking practice, real-time correction, and the social push to actually produce the language. Each patches the other's weakness.
For a beginner without a classroom, the lesson still holds — you just supply the "other half" yourself with real reading, listening, and eventually conversation practice, rather than a formal class. If you're weighing whether an app can stand in for structured teaching, our deeper comparison in is a language app as good as a class walks through exactly what you gain and lose. The framing that actually helps a beginner isn't "which one wins" — it's "what does each one do, and how do I combine them?"
Why some old "it doesn't work" claims were unfair
Before you trust any strong statement about Duolingo — for or against — it's worth knowing that a lot of older claims came from weak studies. Researchers themselves flag that early work often had tiny sample sizes, no proper before-and-after testing with reliable measures, and little control over what learners actually did. A study that measures a handful of people with no real baseline can "prove" almost anything, which is how both breathless praise and blanket dismissal got airtime.
The more recent, better-designed research — larger groups, real pretests and posttests, direct comparison conditions — has generally landed on the same nuanced result: Duolingo is effective for foundational reading, listening, vocabulary, and grammar, and not a standalone route to speaking fluency. That's less exciting than "this app is magic" or "this app is useless," but it's what the good evidence supports. Being skeptical of confident one-liners in either direction is itself a useful language-learning habit, something we touch on in common myths about learning a language.
How to use Duolingo based on the evidence
Put the research together and a clear playbook falls out. First, use Duolingo for the foundation and trust it there. For your first weeks and months, it's a legitimate, evidence-backed way to meet your first words, absorb basic grammar, and build daily momentum — you don't need to feel guilty about starting here or worry it's a waste of time. The studies say it isn't.
Second, add real input early, don't wait to "finish." Because the app's proven gains sit in comprehension and its gap sits in speaking and advanced fluency, start bringing in real language as soon as simple sentences feel familiar — often within the first month or two. Read content matched to your level so you meet lots of understandable language at volume (an especially efficient next step, and the whole idea behind learning through immersion), listen to audio you can mostly follow, and — when you're ready — start speaking in small, imperfect ways. Keep Duolingo as a warm-up if you like it; just let real reading, listening, and conversation become the engine. Used this way, the answer to "does Duolingo actually work?" is a confident yes — for what it's for, and you'll know exactly where to go next.
FAQ
Does Duolingo actually work for learning a language?
Yes, within limits — and this is now backed by peer-reviewed research, not just marketing. Controlled studies have found that beginners who learn on Duolingo reach reading, listening, vocabulary, and grammar levels comparable to students in a traditional classroom over similar periods. In one Cambridge-published study, beginner French learners using Duolingo improved by a similar amount to a classroom-only group and a classroom-plus-Duolingo group across nearly all measures. So the honest answer is: for building the foundations — your first several hundred to few thousand words, basic sentence patterns, and reading and listening comprehension of simple material — Duolingo genuinely works, and works about as well as a beginner class for those specific skills. What the same research does not show is Duolingo carrying you to confident, spontaneous speaking or to advanced fluency. It's a real, evidence-backed way to build the base, not a complete path to the finish line. The sensible read: trust it for the early stages, and plan to add real reading, listening, and speaking practice as you progress.
Is Duolingo as good as a classroom?
For beginner-level reading, listening, vocabulary, and grammar, the research says it's roughly comparable — which surprises a lot of people. A study of elementary Spanish learners found Duolingo instruction was about as effective as classroom instruction for grammar and vocabulary after 40 minutes a week for 12 weeks, and a separate comparison found Duolingo beginners in Spanish and French reached reading and listening skills comparable to university students studying those languages over four semesters. That's a meaningful result: an app you use on the bus performing on par with formal instruction for foundational skills. But 'as good as a classroom' has a big asterisk. A classroom (or a tutor) gives you live speaking practice, real-time correction, and the social pressure that pushes you to actually produce the language — things Duolingo's exercises only imitate. So Duolingo can match a class on the input-and-knowledge side while lagging on the speaking-and-interaction side. The strongest approach in the research was often classroom plus Duolingo, which hints at the real answer: they're complements, not rivals.
Why do people say Duolingo doesn't work?
Usually because they're measuring it against the wrong finish line. The common complaint — 'I did a 500-day streak and still can't hold a conversation' — is real, but it's not evidence the app does nothing. It's evidence that the app is good at one thing (building recognition, vocabulary, and comprehension of simple language) and weak at another (producing fluent, spontaneous speech). The research mirrors this: Duolingo shows solid, measurable gains in reading, listening, vocabulary, and grammar, but studies don't show it delivering confident conversational speaking on its own. So people who expected an app to make them a talker feel let down, while the app was quietly doing exactly what it's good at. There's also a fairness point researchers raise: many older 'Duolingo doesn't work' claims came from weak studies with tiny samples and no proper before-and-after testing. Newer, better-designed research has generally found Duolingo effective for the foundations. The accurate statement isn't 'it doesn't work' — it's 'it works for the base, and you have to add speaking practice yourself.'
How long should you use Duolingo before switching to real content?
There's no single magic number, but a useful rule of thumb is: start bringing in real reading and listening as soon as you have a small base of a few hundred words and can recognise the basic sentence patterns — often within the first several weeks to a couple of months, not after you 'finish the tree.' The research supports using Duolingo for the foundation, and it's genuinely good at that stage. But because the studies show its gains concentrate in early reading, listening, vocabulary, and grammar — and not in advanced fluency or speaking — the smart move is to treat it as a warm-up that overlaps with real language rather than a phase you complete first. In practice: keep doing Duolingo if you enjoy the habit, but once simple sentences feel familiar, start reading content matched to your level and listening to audio you can mostly follow. Reading in particular is an efficient next engine because it exposes you to lots of understandable language at volume. Don't wait for a finish line the app never really gives you.
What does Duolingo NOT teach you well?
Speaking spontaneously, and everything at the higher end of fluency. Across the research, Duolingo's measurable strengths are receptive and foundational: recognising and understanding words, following simple written and spoken sentences, and picking up basic grammar patterns. What it doesn't reliably build is the ability to hold an unscripted conversation, produce complex sentences under real-time pressure, or handle the nuance, idiom, and cultural context that separate 'can pass a beginner test' from 'can actually communicate.' Its exercises are also narrow: tapping word tiles and repeating set phrases isn't the same as composing your own thoughts or reacting to a real person who says something unexpected. Even newer AI conversation features are limited next to talking with an actual speaker. None of this makes the app useless — it makes it a foundation tool. The gaps it leaves (real speaking, advanced comprehension, spontaneous production) are exactly the things you build through conversation practice, extensive reading and listening, and eventually real interaction. Use Duolingo for what it proves it can do, and cover the rest deliberately.