Does speed reading work? What the research actually says
Speed reading courses claim you can read three times faster without comprehension loss. The research says otherwise. Here is an honest look at what RSVP and speed reading techniques actually do — and what they do not.
The real baseline: 238 words per minute
Most speed reading marketing compares your "after" speed to a baseline of 300 WPM — a figure from 1960s studies with small, unrepresentative samples. A 2019 meta-analysis by Marc Brysbaert of Ghent University re-examined the evidence at scale: 190 studies, 18,573 participants. The actual average for silent reading is 238 WPM for non-fiction and 260 WPM for fiction.
This matters for evaluating speed reading claims. If your real baseline is 238 WPM, an app or course promising "3× your normal speed" is promising 714 WPM. The research consensus does not support reading at that speed with comprehension intact.
What research says about speed reading claims
The most thorough review of speed reading research is Rayner et al. (2016), So Much to Read, So Little Time, published in Psychological Science in the Public Interest. Its conclusions are direct: there is a speed-accuracy tradeoff. Comprehension degrades measurably above about 400 WPM. Claims of reading at 600–900 WPM without comprehension loss are not supported by the evidence — those are the paper's own words.
Speed reading courses (as distinct from RSVP) primarily increase skimming: you cover text faster by taking in less. The feeling of reading faster is real; the retention is not always there. RSVP is a different mechanism — it removes eye movement overhead rather than mental processing — but it is subject to the same ceiling: above 400 WPM for most readers, comprehension falls.
What is real: the eye movement overhead. Rayner's work shows saccadic eye movement accounts for roughly 30% of normal reading time. RSVP genuinely eliminates that cost.
The real costs of reading faster
Speed has costs that the marketing does not mention.
Parafoveal preview — in normal reading, the eye pre-processes the next 1–2 words before they enter the focal point. RSVP eliminates this entirely. Your brain gets no head start on the next word; each word must be processed cold.
Regression — skilled readers go back on around 10–15% of words (Rayner et al. 2016). Not from bad habits: from comprehension failure that needs repair. RSVP makes regressions impossible without pausing the entire stream.
Masking — in continuous RSVP, each new word visually masks the previous one. A 2016 study by Primativo et al. published in PLOS One found this masking cost is approximately 180 WPM compared to unmasked conditions.
Working memory — the faster the word stream, the harder it is to hold earlier context. Sentences that span multiple clauses become more difficult to integrate at high speed.
So does it work at all?
Yes, with honest expectations.
For narrative or familiar text at 250–400 WPM, RSVP comprehension is broadly comparable to normal reading. The technique eliminates real overhead (eye movement), creates a steady focal point that reduces mind-wandering, and produces a genuine speed increase in the range the research supports.
It is not a shortcut to reading three times faster without losing anything. Speed reading courses that promise 600–1,000 WPM with full comprehension are overstating what the evidence shows.
The most robustly supported application is accessibility. Rubin and Turano (1992) showed that removing the need for eye movement across the page dramatically helps readers who cannot scan well. Aquilante et al. (2001) found that low-vision patients with age-related macular degeneration read 33% faster with duration-adaptive RSVP (words displayed longer when longer) than with constant-duration RSVP.
RSVP apps vs. speed reading courses
Speed reading courses — from Evelyn Wood to modern app-based training — focus mainly on two things: reducing subvocalization (the inner voice that reads each word) and increasing skimming. The underlying claim is that subvocalization and regression are bad habits to eliminate. The research does not support this framing: subvocalization is part of comprehension, and regressions are repair mechanisms, not bad habits.
RSVP is a different mechanism. It does not try to eliminate subvocalization; it removes the physical cost of eye movement. At moderate speeds, this is a legitimate and measurable gain. It is more honest than courses promising four-digit WPM figures.
What Leo does: a fixed focal point, punctuation-aware rhythm, and a synchronized full-chapter view below the RSVP word so you can regress when you need to — which is the single most important thing a well-designed RSVP reader can offer. Speed is a dial, not a target. The goal is reading your books, not passing a WPM test.
Want to try the honest version of RSVP? Join the waitlist, the first people signed up will be the first to read with Leo.
References
This article is based on the following peer-reviewed sources, which you can read at the original links:
- Brysbaert, M. (2019). How many words do we read per minute? A review and meta-analysis of reading rate. Journal of Memory and Language, 109, 104047.
- Rayner, K., Schotter, E. R., Masson, M. E. J., Potter, M. C. and Treiman, R. (2016). So Much to Read, So Little Time: How Do We Read, and Can Speed Reading Help? Psychological Science in the Public Interest, 17(1), 4–34.
- Primativo, S., Stellato, M., Arduino, L. S., Curcio, C. A. and Miriam, D. C. (2016). Aging effects on the reading of single words in RSVP: the role of presentation rate and masking. PLOS One, 11(4), e0153786.
- Rubin, G. S. and Turano, K. (1992). Reading without saccadic eye movements. Vision Research, 32(5), 895–902.
- Aquilante, K., Yager, D., Morris, R. A. and Khmelnitsky, F. (2001). Low-vision patients with age-related maculopathy read RSVP faster when word duration varies according to word length. Optometry and Vision Science, 78(5).