The Mozart Effect and the Long Search for Smarter Kids

Thirty-six college students, ten minutes of Mozart and a paper-folding puzzle: that was the whole experiment behind one of the most durable parenting beliefs of the last thirty years.

A one-page letter and a national craze

The letter appeared in Nature in October 1993. Its authors, Frances Rauscher, Gordon Shaw and Katherine Ky at the University of California, Irvine, had asked students to sit through one of three conditions before a set of spatial reasoning tasks: ten minutes of Mozart's Sonata for Two Pianos in D major (K. 448), ten minutes of relaxation instructions, or ten minutes of silence. After the Mozart, the students scored somewhat higher on the spatial tasks. The authors translated the difference into the equivalent of roughly eight or nine points on an IQ scale, and they noted that it lasted only about 10 to 15 minutes.

Nobody in the study was a child. Nobody took a music lesson, and nobody's general intelligence was measured over time.

None of that slowed the idea down. Newspapers shortened the finding to "Mozart makes you smarter", and within a few years the phrase "the Mozart effect" had become a book title, a trademark and a shelf of products for infants. In 1998 Zell Miller, then governor of Georgia, proposed spending state money so that every newborn's family would go home from hospital with a classical music recording, and Florida passed a law the same year asking state-funded child care centres to play classical music daily. What had been a short-lived bump in a puzzle score among adults was now public policy for babies.

The replications, and a simpler explanation

Other labs tried to reproduce the result, with mixed and mostly underwhelming outcomes. In 1999 Nature published a pair of critical reports under the heading "Prelude or requiem for the 'Mozart effect'?" One of them, a meta-analysis by Christopher Chabris, pooled the available studies and concluded that whatever advantage existed was small, and smaller than the original letter implied.

Glenn Schellenberg, a psychologist at the University of Toronto, offered an explanation that did not require anything special about Mozart. Music that a listener enjoys can briefly lift arousal and mood, and a person who feels alert and upbeat tends to do a little better on a demanding task. In a 1999 study with Kristin Nantais, participants who heard a short story by Stephen King did as well afterwards as those who heard Mozart, provided they preferred the story. On this reading, the "effect" was real but ordinary: anything pleasant and stimulating would do.

The broadest look at the listening question came in 2010, when Jakob Pietschnig, Martin Voracek and Anton Formann published a meta-analysis in the journal Intelligence covering roughly 40 studies and several thousand participants. Their conclusion, signalled in the paper's title ("Mozart effect, Shmozart effect"), was that there was little evidence of a specific benefit from listening to Mozart.

By then the question had already shifted. If passive listening did little, perhaps actually learning to play was the thing that mattered. Lessons involve years of practice, memorising, reading notation and sustained attention, so a spillover was at least plausible.

Schellenberg tested it directly. His 2004 study in Psychological Science is still one of the most cited experiments in the field, largely because the children were randomly assigned rather than self-selected. About 140 six-year-olds in Toronto received a year of keyboard lessons, voice lessons, drama lessons, or no lessons (the no-lesson group was offered lessons the following year). Every group's IQ rose over the year, as children's scores typically do once they start school. The two music groups rose a little more, by a few IQ points on average. Schellenberg described the difference as small. The drama group, unexpectedly, showed the clearest improvement in social behaviour as rated by parents.

Later work by Schellenberg added a caution: in ordinary life, the children who take lessons for years tend to differ from those who do not, in family income, parental education, personality and persistence. Those differences can make lessons look more powerful in correlational data than they are.

What the large reviews found

Psychologists have a name for the hoped-for spillover: far transfer, meaning that training in one area improves a skill quite unlike it, such as maths or working memory. Giovanni Sala and Fernand Gobet have examined it repeatedly. Their 2017 meta-analysis of music training in children and young adolescents, published in Educational Research Review, found a small overall benefit on cognitive and academic measures, and found that the better designed the study, the smaller the benefit. In 2020 they published a larger multilevel meta-analysis in Memory & Cognition. Once they took into account whether children were randomly assigned and whether the comparison group did some other engaging activity, the effect of music training on cognitive and academic skills was close to zero.

Two randomized trials of preschool music classes, run by Samuel Mehr and colleagues at Harvard and published in PLOS ONE in 2013, pointed the same way: no reliable cognitive advantage over children who took part in comparison activities.

Not everyone reads the evidence so bleakly. In a 2022 paper, Emmanuel Bigand and Barbara Tillmann re-analysed the data and argued that modest benefits remain once certain studies are treated differently. The disagreement is real, but it is now about whether the effect is small or very small, not whether lessons are a route to a higher IQ.

Meanwhile, work closer to music itself has held up better. Nina Kraus's Auditory Neuroscience Laboratory at Northwestern University has studied how training shapes the brain's response to sound, including speech heard against background noise. In one community program in Los Angeles, her team reported measurable changes only after children had taken part for around two years, which says something about how gradual these benefits are.

Pulling the threads together, these are the outcomes lessons reliably build:

• musical skills: pitch, rhythm, reading notation, and the motor control a particular instrument needs
• finer listening, including some evidence for processing sound and speech more precisely after sustained training
• the habit of practising something difficult and seeing measurable progress
• membership of a group, for children in choirs, bands and orchestras

That list is enough to justify lessons, and it changes how to choose them. If the main payoff is musical and personal, the instrument and teacher a child will stick with matter far more than any claimed brain benefit.

Age shapes the options. Under about five, group classes built on singing, movement and rhythm suit most children; some Suzuki teachers start violin or piano around four, but only with a parent who attends lessons and supervises short daily practice. Piano is a common starting point from around five to seven, since it needs no particular hand size, and 30-minute lessons suit that age. Fractional-size string instruments and smaller guitars become realistic from about seven. Wind and brass instruments generally wait until a child has adult front teeth and enough breath control, which is one reason many US schools start band and orchestra between 4th and 6th grade.

Cost varies widely by city. In many US areas in 2026, a weekly 30-minute private lesson commonly runs somewhere around $30 to $60, and school programs and group classes usually cost less per session. Rent the instrument before buying one; rent-to-own plans soften the loss if interest fades or a child outgrows a fractional size.

Before paying for a term, book a single trial lesson and watch how the teacher speaks to your child when they get something wrong. Then ask the teacher what a typical student can play after one year. A specific musical answer is a good sign; a promise of better grades or higher IQ is a reason to keep looking.