Original research

Do Subtitles Help You Learn a Language? What the Research Says

Watching with subtitles is a habit most of us already have. The question is whether it teaches you anything, and which kind of subtitles works best.

By Lachlan McRitchie, Founder, LingoBinge

Updated July 6, 2026

Yes, watching with subtitles helps you learn a language, and the research is fairly clear about it: a meta-analysis of 18 studies found that same-language captions produce a large improvement in both listening comprehension and vocabulary. The catch is which subtitles you use, since captions in the language you are learning help far more than subtitles in your own. This study gathers what the evidence actually shows about subtitles, comprehension, and how many words you pick up from watching.

Key takeaways

  • A meta-analysis of 18 studies found that same-language subtitles produce a large improvement in both listening comprehension and vocabulary learning.
  • In one experiment, an hour of TV with target-language subtitles improved listening by about 17%, while native-language subtitles gave almost no listening gain.
  • Watching a single documentary taught viewers roughly 14% of the words in it they did not already know, just from viewing.
18 studies
pooled in a meta-analysis showed a large positive effect of same-language subtitles on L2 listening and vocabulary

Montero Perez et al., 2013

+17%
listening improvement from an hour of TV with target-language subtitles, versus near zero from native-language ones

Birulés-Muntané and Soto-Faraco, 2016

14%
of the unfamiliar words in a single documentary picked up just from watching it

Peters and Webb, 2018

80%
of 18 to 24-year-olds already watch with subtitles at least some of the time

Stagetext / YouGov, 2021

Subtitles genuinely help, and the evidence is strong

The clearest verdict comes from a 2013 meta-analysis that pooled 18 separate studies of captioned video. It found that same-language captions, the subtitles in the language you are learning, produce a large positive effect on both listening comprehension and vocabulary learning. A large effect in this research means the benefit is not a rounding error; it is one of the more robust findings in the field.

Later work filled in the how. Learners pick up vocabulary incidentally, just from watching, with no intention to study. In one experiment, viewers learned about 14% of the words in a single documentary that they had not known before, simply by watching it once. The effect is real, if modest per sitting, and it compounds the more you watch.

large effect
of captions on both listening and vocabulary, the headline finding of the meta-analysis

Montero Perez et al., 2013

Which subtitles matter: your target language beats your own

Not all subtitles are equal, and the difference is large. In a clean experiment, intermediate learners watched an hour of TV in one of three ways: with subtitles in the language they were learning, with subtitles in their native language, or with none. The learning-language subtitles improved listening by about 17%, clearly the best. No subtitles gave a smaller gain of around 7%, and native-language subtitles produced almost no listening improvement at all.

The reason is intuitive. Subtitles in your own language let you read the plot and tune the foreign audio out, so your ear never has to work. Subtitles in the language you are learning keep your eyes and ears on the same words, linking sound to spelling to meaning. One honest caveat: that study measured listening, not vocabulary, where the picture is more mixed, but for training your ear the direction is clear.

Listening improvement after one hour of TV, by subtitle type (Birulés-Muntané and Soto-Faraco, 2016).
CategoryListening improvement
Native-language subtitles
~0%
No subtitles
+7%
Learning-language subtitles
+17%

Bingeing one show recycles the words you need

There is a reason to stick with one series rather than hop between many. Analysing a corpus of 88 hours of television, researchers found that knowing about 3,000 word families covers roughly 95% of the words on TV. More usefully for a binge-watcher, lower-frequency words recur far more within a single programme than across unrelated ones. A show reuses its own vocabulary, so the words you look up in episode one come back in episode five.

That recurrence is what turns watching into learning. Frequency of encounter is one of the strongest predictors of whether a word sticks, so a series that keeps recycling its vocabulary is doing the spaced repetition for you. It is the argument for finishing the show, not just starting it.

3,000
word families cover about 95% of the words on television, and a single series keeps reusing them

Webb and Rodgers, 2009

Most people already watch this way

The habit is already there. Surveys find that about half of Americans watch with subtitles most of the time, and among younger viewers it is far higher: roughly 70% of Gen Z are frequent subtitle users, and 80% of 18 to 24-year-olds in the UK watch with subtitles at least some of the time. Most of that is by choice, not hearing need.

That is the opportunity. If most people already read subtitles while they watch, the small step from entertainment to learning is not adding a new habit but changing which language the subtitles are in, and pitching them at a level you can actually follow. The research says the habit works; the trick is aiming it at the language you want to learn.

~50%
of Americans watch with subtitles most of the time, rising to about 70% of Gen Z

Preply, 2023

LingoBinge is built on this idea: it turns the shows you already watch into understandable practice, swapping the highest-frequency words first and saving them for spaced review.

Methodology and caveats

The headline effect comes from Montero Perez, Van Den Noortgate and Desmet (2013), a meta-analysis of 18 captioned-video studies that reported a large effect on both listening comprehension and vocabulary. The authors report a bias-corrected effect size (Hedges' g); the open abstract confirms the effect is large without printing the exact decimal, so we describe it as a large effect rather than quoting a specific number. Effects were moderated by test type.

The subtitle-type comparison is Birulés-Muntané and Soto-Faraco (2016), an experiment with intermediate Spanish learners of English watching one hour of TV. It measured listening improvement, where target-language subtitles clearly won; it found no reliable vocabulary gain in any condition, so we cite it only for listening. The single-documentary figure of about 14% is from Peters and Webb (2018), and gains vary widely with the test, the learner, and how often a word recurs, so any per-season total would be an extrapolation, not a measured result.

Coverage and recurrence figures are from Webb and Rodgers (2009), measured on an 88-hour television corpus in word families. Subtitle-usage figures are consumer surveys (Preply 2023; Stagetext and YouGov 2021), which are self-reported and not peer-reviewed, so we treat them as cultural context rather than learning evidence. We deliberately do not repeat the widely-shared claim that 96% of Gen Z use subtitles, which we could not trace to the underlying survey.

Sources

  1. Montero Perez, Van Den Noortgate and Desmet, Captioned video for L2 listening and vocabulary learning: a meta-analysis, System (2013)
  2. Birulés-Muntané and Soto-Faraco, Watching subtitled films can help learning foreign languages, PLOS ONE (2016)
  3. Peters and Webb, Incidental vocabulary acquisition through viewing L2 television, Studies in Second Language Acquisition (2018)
  4. Webb and Rodgers, Vocabulary demands of television programs, Language Learning (2009)
  5. Rodgers and Webb, Incidental vocabulary learning through viewing television, ITL International Journal of Applied Linguistics (2020)
  6. Preply, America's subtitles use survey (2023)
  7. Stagetext and YouGov, Subtitles use among young viewers (2021)

This study and its data are free to cite and reuse under a Creative Commons Attribution 4.0 licence. Please link back to this page.

Frequently asked questions

Do subtitles help you learn a language?
Yes. A meta-analysis of 18 studies found that subtitles in the language you are learning produce a large improvement in both listening comprehension and vocabulary. The key is using captions in your target language, not your native one, and pitching the content at a level you can mostly follow.
Are subtitles in your target language better than in your own?
For training your ear, clearly yes. In one experiment an hour of TV with target-language subtitles improved listening by about 17%, while native-language subtitles gave almost no gain, because they let you read the plot and tune the foreign audio out. Target-language captions keep sound, spelling, and meaning together.
How many words can you learn from watching TV with subtitles?
It varies, but one study found viewers picked up about 14% of the unfamiliar words in a single documentary just from watching. Gains are modest per sitting and depend on how often a word recurs, which is why finishing a series, where vocabulary repeats, works better than hopping between shows.

More research and references