Does Music Help You Focus? From the Mozart Effect to Lo-Fi

In the summer of 1998, every baby born in a Georgia hospital went home with a compact disc. The governor, Zell Miller, had wanted the state to pay for it; the legislature balked, so Sony donated 100,000 copies, enough for a year of births. A note from the governor was tucked into the sleeve: play it often, and your little one will get off to a smart start. Asked by a reporter for the science, Miller explained that a brain has two lobes and that music builds a bundle of nerves between them (Baltimore Sun, 1998).

Twenty-eight years later the disc has become a YouTube stream of a cartoon girl doing homework to lo-fi beats, and the promise has migrated from babies to anyone with a laptop: put the right sound on and you will get more done. Traced back, it still rests on one page of a journal from October 1993. I went and read that page.

One page, 36 students, ten minutes

The paper is a letter in Nature by Frances Rauscher, Gordon Shaw and Catherine Ky of the Center for the Neurobiology of Learning and Memory at the University of California, Irvine (Rauscher, Shaw & Ky, 1993). Nature keeps it behind a paywall; I read a scan hosted by Rauscher’s later university, so the numbers below come from the original page.

Thirty-six college students each sat through three ten-minute conditions: Mozart’s Sonata for Two Pianos in D major, a tape of relaxation instructions, and silence. After each one they took a different abstract-reasoning task from the Stanford-Binet scale (pattern analysis, matrices, or paper folding and cutting). Standard age scores came out at 57.56 after Mozart, 54.61 after the tape, 54.00 after silence, which the authors translated into spatial IQ equivalents of 119, 111 and 110. That is where the famous “8 to 9 IQ points” comes from.

Two sentences on that page never made it into the CD sleeve. The first: the enhancing effect “does not extend beyond the 10–15-minute period during which subjects were engaged in each spatial task”. The second, in the closing paragraph, is a prediction: “music lacking complexity or which is repetitive may interfere with, rather than enhance, abstract reasoning”. The people who discovered the Mozart effect expected repetitive background music to hurt.

From undergraduates to newborns in four years

In 1997 a music-industry entrepreneur named Don Campbell published a book called The Mozart Effect with Avon Books. Its jacket promised music for “everything from anxiety to cancer” and quoted a hospital director’s claim that half an hour of classical music equals ten milligrams of Valium (Campbell, 1997). None of that is in the Nature letter.

Two psychologists later measured the drift. Adrian Bangerter of the University of Neuchâtel and Chip Heath of Stanford searched the fifty largest American newspapers for every mention of the effect from 1993 onward (Bangerter & Heath, 2004). In 1994, 80 percent of articles mentioned college students, the only people ever tested. By 2000 that was down to around 30 percent. Mentions of babies rose from zero in 1995 to 55 percent in 1999. In their own surveys of 496 people in California and Arizona, more than 80 percent had heard of the effect.

What happened when other labs pressed play

Kenneth Steele at Appalachian State University followed the original team’s published procedure with 125 participants instead of 36 (Steele, Bass & Crook, 1999). Groups heard Mozart, silence, or a deliberately repetitive piece by Philip Glass, then did paper folding and cutting. Every group improved by about the same amount from practice; Mozart against silence came to d = 0.06, next to the d = 0.72 the Irvine lab had reported. Mood did move. The Glass group scored highest on tension and anger, the Mozart group lowest, and Steele’s reading was that if the music does anything, it does it through how people feel.

The same month, Christopher Chabris at Harvard pooled every published comparison he could find, 16 studies and 714 people (Chabris, 1999). Mozart against silence came to d = 0.09, or 1.4 IQ points, not distinguishable from zero (P = 0.26). The gain lived entirely in paper-folding-type tasks and vanished on matrices and digit span.

Then came the experiments that explained the residue. Kristin Nantais and Glenn Schellenberg at the University of Toronto found that Mozart or Schubert beat silence, but when the control condition became a narrated story instead of silence the advantage disappeared: people scored best after whichever they preferred, music or story (Nantais & Schellenberg, 1999). I could not open the full paper; that summary is the published abstract, and Chabris, who did read it, describes the story as a passage from Stephen King. Two years later William Forde Thompson’s group at York University in Toronto played either the Mozart sonata or a slow Albinoni adagio and measured enjoyment, arousal and mood alongside the spatial test (Thompson, Schellenberg & Husain, 2001). Only Mozart helped, and once arousal and mood were held constant statistically, the Mozart effect disappeared.

The largest accounting came in 2010 from Jakob Pietschnig, Martin Voracek and Anton Formann at the University of Vienna, who gathered nearly 40 studies and over 3,000 people, including unpublished work (Pietschnig, Voracek & Formann, 2010). Mozart against no music came to d = 0.37, with a confidence interval from 0.23 to 0.52. Any other music against no music came to d = 0.38. Mozart against other music: d = 0.15. Correcting for publication bias pushed the estimates lower, and the effect was noticeably larger in studies from Rauscher’s own group than from anyone else’s. Their closing line: “there is little evidence left for a specific, performance-enhancing Mozart effect.” Stephen King would have done as well.

The question you typed is a different one

Every study above played the music and then stopped it. The claim in the search box is about music playing while you work. The first serious attempt to add that literature up came from Juliane Kämpfe, Peter Sedlmeier and Frank Renkewitz at Chemnitz University of Technology (Kämpfe, Sedlmeier & Renkewitz, 2011). They found 97 usable studies of adults doing something with background music on, from shopping to reading.

Pooled together, the effect on behaviour was a null. A null from studies pointing in opposite directions is an average of a plus and a minus, and when they split the pile by task the two came apart. Eight studies of reading with music on versus off were negative in every single case (weighted r = −0.11, 680 people): a small number, but the wrong sign for a focus aid and the most consistent one in the set. Memory came out slightly negative, emotional reactions positive, sports performance positive (r = 0.15), and the tempo of the music set the tempo of whatever people were doing.

Why words in the room get into words on the page

In 1982 Pierre Salamé and Alan Baddeley showed that people remembering a short list of digits on a screen did worse when speech they had been told to ignore played in the background, and that the damage tracked how much the unattended words sounded like the digits, not what they meant (Salamé & Baddeley, 1982). Their explanation was a short-term store for speech-like sound that you cannot close. In 1989 they ran the same test with music (Salamé & Baddeley, 1989). Vocal music disrupted recall more than instrumental in both experiments, and with well-practised subjects instrumental music was not reliably worse than silence.

Nick Perham and Harriet Currie at Cardiff Metropolitan University ran the version of this that matters for a study playlist (Perham & Currie, 2014). Thirty undergraduates did reading-comprehension passages under four conditions: quiet, instrumental music, lyrical music from their own collection, and thrash metal they had said in advance they disliked. Quiet and instrumental came out best and did not differ. Both lyrical conditions came out worst and did not differ from each other; the songs people loved cost them as much as the songs they hated. Then the questionnaire: participants rated their own liked music as the condition where they had done best.

Thirty people is a small study; the number that carries weight is the pooled one. Martin Vasilev, Julie Kirkby and Bernhard Angele at Bournemouth University ran a Bayesian meta-analysis of 65 studies of reading with sound in the background (Vasilev, Kirkby & Angele, 2018). Background sound cut comprehension by a Hedges’ g of −0.21 overall: speech −0.26, noise −0.17, music −0.19. Lyrical music was more disruptive than non-lyrical by g = −0.19, with a 95 percent probability that the difference is real, and lyrical music did not reliably differ from intelligible speech: a song with words costs a reader about what a talking colleague costs. Their practical sentence: instrumental music while reading “does not affect the comprehension of the text, whereas listening to lyrical music does.”

Setting	Study	What was measured	Result
Mozart before a spatial test	Rauscher et al., 1993 (n = 36)	Stanford-Binet spatial subtest	+8–9 “IQ points”, gone in 10–15 min
Mozart before a spatial test	Chabris, 1999 (16 studies, 714 people)	Mozart vs silence	d = 0.09, 1.4 IQ points, P = 0.26
Mozart before a spatial test	Pietschnig et al., 2010 (~40 studies, 3,000+ people)	Mozart vs silence · any music vs silence · Mozart vs other music	d = 0.37 · 0.38 · 0.15
Music during a task	Kämpfe et al., 2011 (97 studies)	Reading with music on	r = −0.11, negative in all 8 studies
Music during reading	Vasilev et al., 2018 (65 studies)	Comprehension with lyrics vs without	g = −0.19 extra cost for lyrics
Music during four tasks	Souza & Leal Barbosa, 2023 (n = 100–123)	Lyrics vs silence · lo-fi vs silence	d ≈ −0.3 · no credible effect

Where the music does help: crossing out the letter A

Manuel Gonzalez, then at Baruch College, and John Aiello at Rutgers gave 150 undergraduates two tasks (Gonzalez & Aiello, 2019). The simple one was crossing out every word containing the letter A in a list. The complex one was memorising pairs of words and recalling one from the other. Participants worked in silence or with instrumental music, stripped down or layered, soft or loud. Music as a whole hurt the complex task; layered music helped the simple one; and how much external stimulation a person prefers shifted both. The authors’ summary is the honest version of every “it depends” blog post: the effect “depends on the music, the task, and the performer.”

That split matches what Kämpfe’s team saw across 97 studies and what a 2022 review of 95 articles and 154 experiments found by counting the direction of every result: a general detrimental effect on memory and language tasks, lyrics worse than instrumental, difficult tasks hurt where easy ones were not, introverts hurt where extraverts were not, and exactly one positive effect in the entire pile, from instrumental music (Cheah, Wong, Spitzer & Coutinho, 2022).

Lo-fi, tested once

I looked for a meta-analysis of lo-fi specifically and found none in Crossref or Europe PMC as of September 2026; the genre is younger than the review cycle. What exists is one well-powered experiment. Alessandra Souza and Luís Carlos Leal Barbosa at the University of Porto had between 100 and 123 psychology students do verbal memory, visual memory, reading comprehension and arithmetic under three conditions: silence, lo-fi hip-hop, and pop songs with Portuguese lyrics (Souza & Leal Barbosa, 2023). Lyrics cut verbal memory, visual memory and reading by around d = −0.3; the −0.19 on arithmetic did not clear their credibility threshold. Lo-fi did not credibly help or hurt anything, which beats the pop playlist and falls short of the stream’s title. Participants correctly judged that the lyrics were costing them, and sometimes judged the lo-fi to be helping, which the scores did not support.

The two things people mean by “helps me focus”

Read side by side, the blogs and the labs mostly stop disagreeing once you separate two claims that share a sentence. One is about how the work feels. Steele’s Mozart group felt less tense than the silence group, Thompson’s participants were more aroused by the sonata, and Kämpfe’s emotion outcomes were the one category that came out reliably positive; for a task you have been avoiding for three days, that may be the whole game. The other claim is about how much you understood, remembered or produced, and there the pattern since 1982 has not changed: repetitive work gets a little faster with something lively on, anything that runs through language gets a little worse with words in your ears, and instrumental music mostly lands on zero. The loss is small enough to be invisible from the inside, which is why Perham’s participants and Souza’s participants both told the questionnaire the opposite of what the score sheet said.

If your instinct after this is to swap the songs for a waterfall rumble, the trials of brown, pink and white noise for concentration are counted in the piece on brown noise and focus, and the wider question of how attention breaks and comes back is in the attention section, including how long recovery takes after an interruption and the eight-second goldfish statistic that never existed. The full register of claims we have traced is at the myths ledger.

Sources

• The Baltimore Sun (1998). Tuning in to brain development: The music-loving governor of Georgia is providing every newborn in his state with a CD of classical music. The Baltimore Sun, 6 July 1998. Georgia CD programme: Sony donated 100,000 discs after the legislature refused to fund it; Zell Miller's two-lobes explanation to a reporter baltimoresun.com
• Rauscher, F. H., Shaw, G. L., & Ky, C. N (1993). Music and spatial task performance. Nature. 365(6447), 611; N = 36, three 10-min conditions, Stanford-Binet spatial subtests; standard age scores 57.56 / 54.61 / 54.00 (spatial IQ 119 / 111 / 110); effect gone within 10-15 min; authors predict repetitive music may interfere doi:10.1038/365611a0
• Campbell, D. G (1997). The Mozart Effect: Tapping the Power of Music to Heal the Body, Strengthen the Mind, and Unlock the Creative Spirit. Avon Books, New York (Internet Archive scan). Jacket copy: music for "everything from anxiety to cancer", half an hour of classical music equals ten milligrams of Valium; the commercial vehicle that carried the claim from students to babies archive.org
• Bangerter, A., & Heath, C (2004). The Mozart effect: Tracking the evolution of a scientific legend. British Journal of Social Psychology. 43(4), 605-623; newspaper coverage 1993 onward: college-student mentions 80% (1994) to ~30% (2000), infant mentions 0% (1995) to 55% (1999); >80% of 496 surveyed had heard of the effect doi:10.1348/0144666042565353
• Steele, K. M., Bass, K. E., & Crook, M. D (1999). The Mystery of the Mozart Effect: Failure to Replicate. Psychological Science. 10(4), 366-369; N = 125, Mozart vs silence vs Philip Glass, paper folding and cutting; Mozart vs silence d = 0.06 against the original d = 0.72; Glass group highest on tension and anger doi:10.1111/1467-9280.00169
• Chabris, C. F (1999). Prelude or requiem for the 'Mozart effect'?. Nature. 400(6747), 826-827; meta-analysis of 16 studies, 714 people; Mozart vs silence d = 0.09 (1.4 IQ points), P = 0.26; effect confined to paper-folding-type tasks doi:10.1038/23608
• Nantais, K. M., & Schellenberg, E. G (1999). The Mozart Effect: An Artifact of Preference. Psychological Science. 10(4), 370-373; Mozart or Schubert beat silence, but not a narrated story; people scored best after whichever they preferred (abstract only; full text not opened) doi:10.1111/1467-9280.00170
• Thompson, W. F., Schellenberg, E. G., & Husain, G (2001). Arousal, Mood, and The Mozart Effect. Psychological Science. 12(3), 248-251; Mozart sonata vs Albinoni adagio with enjoyment, arousal and mood measured; the Mozart advantage disappeared once arousal and mood were controlled doi:10.1111/1467-9280.00345
• Pietschnig, J., Voracek, M., & Formann, A. K (2010). Mozart effect-Shmozart effect: A meta-analysis. Intelligence. 38(3), 314-323; ~40 studies, 3,000+ people incl. unpublished; Mozart vs no music d = 0.37 [0.23, 0.52], other music vs no music d = 0.38, Mozart vs other music d = 0.15; publication bias; larger effects in Rauscher's group doi:10.1016/j.intell.2010.03.001
• Kämpfe, J., Sedlmeier, P., & Renkewitz, F (2011). The impact of background music on adult listeners: A meta-analysis. Psychology of Music. 39(4), 424-448 (online 2010); 97 studies; overall null on behaviour; reading negative in all 8 studies (r = -0.11, n = 680), memory slightly negative, emotion positive, sports r = 0.15, tempo effects doi:10.1177/0305735610376261
• Salamé, P., & Baddeley, A (1982). Disruption of short-term memory by unattended speech: Implications for the structure of working memory. Journal of Verbal Learning and Verbal Behavior. 21(2), 150-164; unattended speech disrupts digit recall; disruption tracks phonological similarity, not meaning; phonological store account doi:10.1016/S0022-5371(82)90521-7
• Salamé, P., & Baddeley, A (1989). Effects of Background Music on Phonological Short-Term Memory. The Quarterly Journal of Experimental Psychology Section A. 41(1), 107-122; vocal music disrupted recall more than instrumental in both experiments; with practised subjects instrumental music not reliably worse than silence doi:10.1080/14640748908402355
• Perham, N., & Currie, H (2014). Does listening to preferred music improve reading comprehension performance?. Applied Cognitive Psychology. 28(2), 279-284; N = 30; quiet and instrumental best and equal; liked and disliked lyrical music worst and equal; participants rated liked music as their best condition doi:10.1002/acp.2994
• Vasilev, M. R., Kirkby, J. A., & Angele, B (2018). Auditory Distraction During Reading: A Bayesian Meta-Analysis of a Continuing Controversy. Perspectives on Psychological Science. 13(5), 567-597; 65 studies; background sound g = -0.21 (speech -0.26, noise -0.17, music -0.19); lyrical vs non-lyrical g = -0.19 with 95% probability; lyrical music not reliably different from intelligible speech doi:10.1177/1745691617747398
• Gonzalez, M. F., & Aiello, J. R (2019). More than meets the ear: Investigating how music affects cognitive task performance. Journal of Experimental Psychology: Applied. 25(3), 431-444; N = 150; simple (letter A cancellation) vs complex (paired-associate recall) tasks, instrumental music varied in complexity and volume; music hurt the complex task, layered music helped the simple one; preference for external stimulation moderated doi:10.1037/xap0000202
• Cheah, Y., Wong, H. K., Spitzer, M., & Coutinho, E (2022). Background Music and Cognitive Task Performance: A Systematic Review of Task, Music, and Population Impact. Music & Science. Vol. 5; 95 articles, 154 experiments; general detrimental effect on memory and language tasks, lyrics worse than instrumental, difficult tasks and introverts hurt, one positive effect (instrumental) doi:10.1177/20592043221134392
• Souza, A. S., & Leal Barbosa, L. C (2023). Should We Turn off the Music? Music with Lyrics Interferes with Cognitive Tasks. Journal of Cognition. 6(1); N = 100-123 per task; silence vs lo-fi hip-hop vs Portuguese-lyric pop across verbal memory, visual memory, reading, arithmetic; lyrics d ~ -0.3 on three tasks (-0.19 arithmetic not credible); lo-fi no credible effect doi:10.5334/joc.273