It has been suggested that vaccination, particularly with the mumps, measles, and rubella (MMR) vaccine, may be related to the development of autism. One of the arguments for a possible association is that the prevalence of autism has been increasing at the same time that infant vaccination coverage has increased. Although the prevalence of autism and similar disorders appears to have increased recently, whether this is an actual increase or due to increased recognition and changes in diagnostic criteria is not clear. The clinical evidence in support of an association derives from a report of 12 patients with inflammatory bowel conditions and regressive developmental disorders, mostly autism.1 The authors' speculation that MMR vaccine may cause autism through a persistent bowel infection has generated much interest because it provides a possible biologic mechanism for a causal association. A subsequent epidemiologic study in the United Kingdom, however, did not find an association between MMR vaccination and the onset of autism.2 Several different analyses were performed, all of which showed no association. One of the findings of the study was that although autism prevalence had increased in the study population, the increase started before the introduction of routine MMR vaccination in Britain in 1987, and there was not an abrupt increase in prevalence after the vaccine's introduction. The interpretation of this finding, however, has been debated. The study by Kaye et al provides evidence that the autism trends in the United Kingdom are not related to MMR vaccination. The study was a time trend analysis of a large database of general medical practices in the United Kingdom and compared diagnoses of autism with coverage of MMR vaccine. There was a striking 7-fold increase in the incidence of newly diagnosed cases of autism from 1988 to 1999. Among boys, the incidence of autism increased 4-fold for those born in the years from 1988 to 1993. For the same birth cohorts, the coverage of MMR vaccine was stable at more than 95%. Thus, there was no correlation between the large increase in autism diagnoses and MMR coverage. Studies in Sweden and California have also not found any temporal correlations between MMR coverage and autism trends.3,4 The epidemiologic findings are consistent with current understanding of the pathogenesis of autism, which has a strong genetic component and whose neurologic defects probably occur early in embryonic development. It seems unlikely that a vaccine that is given after birth could cause autism. A few cases of autism may have onset after 1 year of age (regressive autism), but the 1 epidemiologic study that included such cases did not find an association with MMR vaccination.2 Currently, the weight of the available epidemiologic and related evidence does not support a causal association between MMR vaccine—or any other vaccine or vaccine constituent—and autism. The possibility of an idiosyncratic reaction in certain susceptible persons cannot be ruled out, but such occurrences would have to be rare and have not been detectable in large epidemiologic studies. In general, parents and physicians should be reassured that MMR vaccine does not cause autism.