The so-called "Mozart effect" refers to two quite different phenomena. The one that has received the most media play concerns the almost magical (and mythical) effect of Mozart's music on intelligence. It is the result of a misrepresentation of the research results. Rauscher, Shaw, and Ky's 1993 study found that 10 minutes of exposure to Mozart's Sonata for Two Pianos in D Major K. 448 temporarily enhanced performance on three spatial reasoning tasks.
The source of the misunderstanding lay in the fact that spatial reasoning is a component of IQ tests, and the researchers reported an increase of some 8 or 9 points in students' IQ scores after listening to the music. The effect lasted some ten to fifteen minutes.
Even in this limited sense, the effect has not been consistently replicated - indeed, it would be fair to say it has more usually failed to be replicated. Moreover, a meta-analysis of studies that have investigated this effect has found that any cognitive improvement "is small and does not reflect any change in IQ or reasoning ability in general, but instead derives entirely from performance on one specific type of cognitive task and has a simple neuropsychological explanation"1.
There does seem to be a case that particular types of music can have an effect on brainwaves - there has been some interesting work done on its possible therapeutic role in reducing epileptic seizures - but the main effect of music seems to be through its effect on arousal.
Most of the research done into the Mozart Effect has continued the example of the original researchers by comparing the effect of listening to Mozart's music with listening to silence or to a relaxation tape. Obviously enough, these various situations would be expected to differentially affect mood and level of arousal (which are known to have a, small and unreliable, effect on cognition). There is evidence that when this effect is controlled for, the Mozart effect (which we may note is also small and unreliable) disappears.
There is however another Mozart effect that promises to be more useful. This is the possibility that formal training in music yields nonmusical benefits. Once again, the media are keen to hypothesize that this effect is on IQ (what is the media's obsession with IQ?). There does however seem to be growing evidence that musical training benefits other faculties - specifically, verbal memory.
BBC radio programme: http://www.bbc.co.uk/radio4/science/mozarteffect.shtml
about the effect of music training from one of the original "Mozart effect" researchers: