What makes reading efficient?
Finishing a document is only part of the job. You may need to remember its argument, spot an exception, compare it with another source or make a decision. An efficient reader gets the required understanding with a sensible use of time. That goal is more useful than making every passage move at the same speed.
Reading brings several processes together: recognizing written words, connecting their meanings, interpreting sentences and relating new information to what you already know. Models of comprehension describe this construction of a coherent message, while reading-systems research places word knowledge within the wider process. [1] [2]
Consider a project report. You might scan the contents to locate costs, read the assumptions closely, then move quickly through a familiar background section. Those are three sensible reading decisions within one document. A single average speed would conceal why you changed pace and whether you found the important limitation.
“What do I need to be able to explain, find or do after reading this?” Decide that first. Then choose a method and a way to check whether it worked.
How the eyes and language work together
Your eyes make brief stops, called fixations, and jumps, called saccades. Information gathered around each stop helps guide what happens next. Eye-controlled display experiments show that useful preview exists, but that different kinds of information are available over different parts of that preview area. Seeing the shape or length of an approaching word is different from understanding several complete lines. [3]
Eye movements also respond to language. A familiar word in an easy sentence is a different processing task from an unfamiliar technical term inside a qualification. The E-Z Reader model is one influential attempt to explain that relationship between word processing and eye behavior. It should be read as a scientific model, with assumptions and limits, rather than a simple instruction to move your eyes in a prescribed pattern. [4]
For a reader, the practical implication is to work on both the material and the method. If an unfamiliar term repeatedly stops you, learning how it is used may be more helpful than trying to force your eyes past it. If the words are familiar but you keep losing the argument, ask what relationship or conclusion you have missed.
Speed, skimming and depth are different choices
A quick first pass can tell you whether an article deserves careful attention. A close reading can reveal whether its conclusion follows from the evidence. Neither approach replaces the other. The mistake is to report both activities as if they were the same kind of reading.
| Your goal | Useful approach | Check success by… |
|---|---|---|
| Locate a fact | Scan headings, terms and likely sections. | Finding the fact and checking its conditions. |
| Get the main argument | Preview, skim and select passages to read closely. | Explaining the central claim and supporting reasons. |
| Learn for later use | Read for connections, retrieve, check and revisit. | Recalling and applying the ideas after a delay. |
| Enjoy a story | Use the pace that preserves the experience you want. | Following the story and enjoying the language. |
Our recommendation is to treat pace as a choice you can revise. Slow down at a turning point, definition or exception. Move on when a section has served its purpose. You do not have to defend a high number when a lower one helps you make a better decision.
What does a reading-speed number tell you?
Brysbaert’s meta-analysis estimated average adult silent reading in English at 238 words per minute for nonfiction and 260 for fiction. Those population averages are context, not an individual target or ceiling. [5]
A test result also depends on the passage, how words are counted, when the timer starts and stops, whether rereading is allowed and what questions follow. Comparing a familiar short story with an unfamiliar policy document tells you little about improvement. Comparing several similar, previously unread passages is much more informative.
Suppose you read 600 words in three minutes and answer eight of ten relevant questions. A later result of two minutes and five correct answers is faster, but it has not demonstrated better performance for a task that needs those missing details. Conversely, the same understanding in less time may be useful. Keep speed and comprehension visible as separate results.
The measurement guide explains fair comparisons, small-test uncertainty and why an app’s presentation rate should not be confused with a measured reading rate.
Understanding now and remembering later
A page can feel clear while it is in front of you. The more demanding test is whether you can explain its point after closing it, then use that information later. Immediate recognition, delayed recall and practical application answer different questions about what you learned.
Experiments with prose learning found that retrieval could improve delayed retention even when additional study produced stronger immediate results. A separate research synthesis found that the effects of spacing depend partly on the delay before the final test. [6] [7]
For study or professional reading, try a modest cycle: read a meaningful section, close it, explain the main point and check what you missed. Return to important material later. Write a question that requires an explanation, rather than only recognizing a phrase. This is our suggested routine, not a promise of a particular speed gain.
Where can training help?
“Reading training” can mean very different activities. A word-recognition drill, vocabulary lesson, guided passage and recall exercise do not have identical goals. An improvement should be described at the level actually tested: better performance on the drill, better comprehension on new passages, or better retention over time.
A synthesis of second-language extensive-reading studies found benefits for reading proficiency in the included learning contexts. That is a different kind of evidence from an exceptional-speed demonstration or a change in an exercise score. [8]
When evaluating a course, look for relevant unfamiliar material, clear comprehension measures and a comparison that can separate training from repeated testing. Our own exercise samples let you explore the format; the fact that an exercise feels engaging does not establish its broader effect. Check the change against the reading you actually want to improve.
The training evidence guide explains the difference between a promising mechanism, a measured task improvement and evidence of useful transfer.
How to read a bold claim
Begin with the outcome hidden behind the headline. Was the reader summarizing a familiar subject, finding selected information or answering detailed questions about unfamiliar text? Was the advertised number a peak, a group average or the display’s setting? A number without those details is difficult to evaluate.
The major review by Rayner and colleagues found no evidence for dramatic reading-rate increases while maintaining excellent comprehension. It distinguishes this claim from useful selective reading. [9]
That conclusion leaves room for worthwhile goals: clearer vocabulary, fewer avoidable restarts, better selection of relevant sections and a more accurate sense of what you understood. It also leaves a responsibility to test those goals honestly. A technique can be useful for one purpose without validating every promise attached to it.
Find the question you want to explore
You can read this section as a course or use it as a reference. Start with the mechanism if you want to understand what a tool changes. Start with measurement if you are comparing results. Start with context if the immediate problem is a long reading list or a demanding document.
- Eye movements and visual processing: what preview, fixation and rereading contribute.
- Comprehension and speed: how gist, detail and inference can diverge.
- Measurement: timing, questions, baselines and comparisons.
- Training: task gains, transfer and study quality.
- Common claims: inner speech, whole-page reading and word streams.
- Techniques: a usable routine before, during and after reading.
- Context: study, work, screens, another language and leisure.
- History: how measurement and marketing developed together.
- Research library: original sources with findings and limitations.
Common questions
Does speed reading work?
Some approaches help with particular tasks, such as selecting relevant information or developing reading proficiency. A claim of much faster full comprehension needs evidence from unfamiliar material and appropriate questions. Define the task before deciding what “works” means.
Is reading slowly always better?
No. A slow pace can still accompany distraction or misunderstanding, while familiar material may be understood efficiently. Use an outcome check—an explanation, a decision or later recall—to see whether the pace serves your purpose.
How should I start measuring progress?
Choose several comparable unfamiliar readings. Record time and comprehension separately, include questions about relationships as well as details, and repeat the comparison over more than one session. Avoid judging improvement from one unusually easy passage.
Sources for this page
These links lead to the original paper, publisher or author archive. For annotations and limitations, visit the research library.
- The role of knowledge in discourse comprehension: A construction-integration modelWalter Kintsch (1988). Theory and model
- Word Knowledge in a Theory of Reading ComprehensionCharles Perfetti and Joseph Stafura (2014). Theoretical synthesis
- The span of the effective stimulus during a fixation in readingGeorge McConkie and Keith Rayner (1975). Experiment
- Toward a model of eye movement control in readingErik Reichle, Alexander Pollatsek, Donald Fisher and Keith Rayner (1998). Computational model
- How many words do we read per minute? A review and meta-analysis of reading rateMarc Brysbaert (2019). Meta-analysis
- Test-enhanced learning: Taking memory tests improves long-term retentionHenry Roediger and Jeffrey Karpicke (2006). Experiments
- Distributed practice in verbal recall tasks: A review and quantitative synthesisNicholas Cepeda, Harold Pashler, Edward Vul, John Wixted and Doug Rohrer (2006). Meta-analysis
- A Meta-Analysis of Extensive Reading ResearchTakayuki Nakanishi (2015). Meta-analysis
- So Much to Read, So Little Time: How Do We Read, and Can Speed Reading Help?Keith Rayner, Elizabeth Schotter, Michael Masson, Mary Potter and Rebecca Treiman (2016). Research review