Does Reading Books Actually Help You Learn a Language?
Yes, and the research is solid, but the effect is smaller and more specific than many fans claim. What the studies found, and how to use it.
Yes. Reading books does help you learn a language, and this is one of the better-studied questions in language learning. It just isn't the miracle some people describe. The effect is real and moderate, and it depends partly on how you go about it.
I care about this question more than most. I'm the developer behind FluenTech. I got stuck at beginner level with Duolingo, and what I really wanted was to read Russian classics. Today I can read Dostoevsky without outside help and understand most of it. But one person's story proves nothing, so here's what the research says.
What the studies found
Researchers call this "extensive reading": reading a lot, mostly for the meaning and enjoyment, rather than picking apart a short text line by line.
The most useful recent summary is a 2025 meta-analysis by Sangers and colleagues, which combined earlier studies. It found positive effects in every area it looked at:
- reading comprehension
- vocabulary
- decoding and reading fluency
- motivation
- writing
- speaking
- general language proficiency
The effects ranged from small to medium, which is the part that often gets left out.
Earlier work points the same way. Nakanishi's 2015 meta-analysis found that learners in extensive reading programs did better than comparison groups, with an average effect of d = 0.46. Jeon and Day found a similar result for reading proficiency across 51 comparisons, d = 0.57.
If you don't read statistics for fun: those numbers mean a real, noticeable advantage, but not a huge one. Think of reading as something that steadily improves many skills at once, rather than a shortcut.
What the research doesn't prove
That reading alone will make you fluent. Speaking improved in these studies too, but practising speaking is still the direct way to get better at speaking. If you want conversations, you need conversations as well.
That translated, side-by-side books work best. I looked, and I couldn't find a meta-analysis that tests this directly. The closest evidence is on glosses, the short word explanations placed beside a text. A 2026 meta-analysis by He and colleagues found they help learners pick up vocabulary while reading (d = 0.94 on average). Results varied a lot between studies, and the analysis only covered tests taken straight afterwards, so it says little about what you remember months later.
I should be upfront here: our apps use line-by-line translation, so I have a stake in this. It's a sensible way to make hard books readable, but it hasn't been proven better than other kinds of support.
How to get more out of reading
Two findings from the 2025 analysis are worth acting on. Effects were larger when learners had a limited choice of texts, and when there was some form of accountability.
1. Choose from a short list. Pick three books you could manage and choose the one you most want to read. Browsing everything ever written in your language is a good way to end up reading nothing.
2. Give yourself a small check. The studies don't point to one best method, so keep it simple: write two sentences about each chapter, answer a few questions, or tell a friend what happened. Anything that makes you notice whether you followed the story.
The rest of this list is practical advice rather than research findings:
3. Start easier than your pride wants. One widely cited estimate from Paul Nation is that reading English unassisted takes 8,000 to 9,000 word families, if you need to know 98% of the words on the page. That threshold has been questioned, and other languages differ, but the point stands: original novels are a big step, so use graded readers or translation support on the way there.
4. Don't look up every word. Look up the ones that stop you following the story, or that keep coming back. The rest can wait.
5. Read regularly and for longer stretches. Ten pages of a story you care about will usually teach you more than one page studied to death, because you meet the same words again in new sentences.
6. Drop books you hate. The whole method relies on you continuing to read. A boring book that you abandon teaches you nothing.
When to start
Earlier than you think, as long as the text matches your level. After you can read the alphabet and know some common words, very simple stories become possible. Classics can come later, and most people will want some support when they get there.
If you'd like to see what that support looks like, you can read the opening pages of White Nights by Dostoevsky, The Metamorphosis by Kafka or Madame Bovary by Flaubert for free in the original, with a translation under every line.
We built our language apps around the same idea: vocabulary and grammar to get you started, graded reading from A1 to B2, then full classics with a translation under each line. They're free to download, though some books and extra content need diamonds you earn while learning, or Premium.
Sources
- Sangers, van der Sande, Welie, Dobber and van Steensel (2025), Learning a Language Through Reading, Educational Psychology Review
- Nakanishi (2015), A Meta-Analysis of Extensive Reading Research, TESOL Quarterly
- Jeon and Day (2015), The effectiveness of core ER principles, Reading in a Foreign Language
- He, Shaari and Ng (2026), The effects of glosses on English L2 incidental vocabulary learning through reading, Frontiers in Language Sciences
- Nation (2006), How Large a Vocabulary Is Needed for Reading and Listening?, The Canadian Modern Language Review