The best-supported way to read faster is unglamorous: read more, so that more words and ideas are familiar when you meet them. Beyond that, most of the time you can save comes from choosing the right kind of reading for each text, not from moving your eyes differently. Realistic gains in true reading speed are modest, and techniques that promise large ones mostly trade away comprehension.

What actually limits your speed

It helps to know where the time goes. Rayner et al. (2016), reviewing the research on reading and speed reading, conclude that the main limit on reading speed is not the eyes. It is how quickly you can recognise words and work out what a sentence means. Just & Carpenter (1980) showed the same thing at the level of individual words: readers spend longer on words they rarely see and at points where meaning has to be put together.

The consequence is that speed follows language skill. A reader who knows the vocabulary and has some background in the subject recognises words faster and needs fewer pauses to make sense of them. This is why the same person reads a familiar topic quickly and an unfamiliar one slowly, a pattern covered in our guide on how text difficulty affects reading speed.

What the evidence supports

Read more, and read more widely

Rayner et al. (2016) identify language skill, including vocabulary, as the main driver of reading speed, and practice at reading as the way it is built. There is no shortcut here, but there is a clear direction:

  • Read regularly, in volume, at a comfortable pace.
  • Read across subjects, so that more vocabulary and more background knowledge become familiar.
  • In a field you need for work or study, read the introductory material first. Later texts in that field get faster because the terms stop being new.

The effect is slow and specific. Reading a lot of economics makes you faster at economics more than at chemistry.

Match your speed to your purpose

Carver (1992) argues that reading is not one process run at different speeds but several distinct processes, which he calls gears. For the college students he studied, he gives typical rates of about 600 WPM for scanning, 450 for skimming, 300 for ordinary reading, 200 for reading to learn and 138 for memorising. These are his estimates for that population, not universal constants, but the structure is useful: the gears differ in what you get out of them, not just in pace.

Many slow readers are not slow at all. They are using a learning gear on text that only needs ordinary reading, or reading every word of something they only needed to scan for one fact. Before you start, ask what you need: one fact, the general argument, or a working understanding you can use later. Then read accordingly.

Preview the structure first

Spend a minute on the title, headings, any summary and the conclusion before reading the body. This is a practical habit rather than a research finding from the sources cited here, but the reasoning fits what is known: comprehension takes less effort when you already know where the text is going, and you can decide which sections deserve full attention.

Reduce interruptions

An interruption costs more than the seconds it takes. You lose your place in the argument and usually re-read a paragraph to get it back. Silence notifications, close the other tabs, and read in blocks long enough to finish a section. As a rule of thumb, an uninterrupted 20 minutes gets more read than 30 minutes that is broken up.

Practise skimming as its own skill

Skimming is worth getting good at, as long as you treat it as a separate activity with a different result. Duggan & Payne (2009) studied skimming under time pressure, and their findings suggest that skimmers tend to pick up the more important ideas from the parts they sample, while losing detail. Rayner et al. (2016) reach a similar view: skimming is effective when the gist is enough.

To practise deliberately:

  1. Read the opening and closing paragraphs in full.
  2. Read the headings and the first sentence of each paragraph.
  3. Stop and read properly wherever something is new, surprising or directly relevant to your question.
  4. Afterwards, state the main point in a sentence. If you cannot, you skimmed too thinly.

Do not use skimming for text where you will be held to the details.

Measure progress with comprehension attached

A speed figure without a comprehension check cannot tell improvement from carelessness. The site’s reading speed test is built around this. You pick one of five original passages, start the timer by clicking Start Reading, read naturally, click Finished Reading and answer four comprehension questions. It reports your raw WPM, your time and your comprehension score.

Say you read a 380-word passage in 1 min 35 sec. That is 240 WPM. Answering all four questions correctly is encouraging evidence for that attempt, though it does not establish complete understanding or a reliable pace for every similar text. If a later attempt shows 290 WPM with two of four correct, do not count the speed increase alone as improvement. It could reflect skimming, passage difficulty or an error on the short quiz. The reasoning behind this is set out in our guide on reading speed versus comprehension.

Test on different passages, at a similar time of day, and compare several results rather than two. Four questions is a coarse check, and single attempts vary. The guide to finding your personal reading speed baseline explains how to build a figure you can plan with.

What does not work as advertised

Rayner et al. (2016) examined the common speed-reading techniques and found little support for them.

  • Eliminating subvocalisation. The inner voice you hear while reading is often described as a brake. The review concludes that inner speech supports comprehension, and that trying to suppress it tends to hurt understanding rather than free up speed.
  • Widening your “eye span”. Drills that claim to train you to take in whole lines or blocks at a glance run into a physical limit. The eyes take in only a small window of text clearly at each fixation, because sharp vision is confined to the centre of the visual field. The review finds no good evidence that practice enlarges this window in a way that speeds reading.
  • One-word-at-a-time (RSVP) apps. These flash words in a fixed position so your eyes do not move. They also remove your ability to look back and to slow down on a hard word, and the review concludes they tend to reduce comprehension of longer texts.
  • Stopping all re-reading. Looking back is part of how readers repair misunderstandings. Training yourself never to do it removes a tool, not a fault.

What these methods usually produce, when they produce anything, is faster skimming. That can be useful, but it is not the same as reading faster with the same understanding.

How much improvement to expect

Be sceptical of any specific promise. The review by Rayner et al. (2016) found that claims of doubling or tripling reading speed with comprehension intact are not supported by evidence. Gains from building vocabulary and knowledge are real but gradual, and they show up mostly in the subjects you read most.

Larger savings can come from decisions about what to read. Consider a 3,000-word article. At 238 WPM, the adult research average for English non-fiction reported by Brysbaert (2019), it takes 12 min 36 sec. At a brisk 300 WPM it takes 10 min, a saving of 2 min 36 sec, and only if the text is easy enough for you to hold that pace. If a one-minute preview shows that a self-contained 1,000-word section answers your question, reading that section at 238 WPM takes about 4 min 12 sec: roughly 5 min 12 sec including the preview. That saves time by narrowing the task, provided the omitted context is not needed.

A plan you can follow

  1. Take the speed test on two or three passages and write down WPM and comprehension for each.
  2. For the next month, read daily, including some material outside your usual subjects.
  3. Before each substantial text, name your purpose and preview the structure.
  4. Use skimming on purpose, for texts where the gist is enough.
  5. Retest on passages you have not read before, and count a result as progress only if comprehension held.

If your speed stays about the same but you are choosing better what to read closely, you are still getting through more. That outcome is the one the research suggests you should expect.