The gene for spinocerebellar ataxia type 1 (SCA1) is a potential candidate gene for schizophrenia because of previous positive linkage findings in this region (6p22-24), and because the reported correlation between SCA1 onset and the number of CAG repeats suggests anticipation. To test the involvement of this gene in the development of schizophrenia, we examined genotypes of the SCA1 CAG repeat polymorphism for 49 Caucasian patients with schizophrenia, and 88 Caucasian controls. We found a significant association between the frequencies of alleles of this gene and schizophrenia (chi 2 = 18.40, df = 8, P = 0.018). Among 13 alleles, one allele (31 trinucleotide repeat) was significantly more frequent in patients with schizophrenia than in controls (chi 2 = 9.57, df = 1, P = 0.002). This association was sustained after applying a Bonferroni correction for multiple testing (P = 0.05/13 = 0.004), and the chi-square results were shown to be robust through Monte Carlo simulation. We observed no allelic association with three flanking microsatellite markers (D6S288, D6S1605, and D6S337), suggesting that our result was not due to population stratification. Further studies of this locus are needed to confirm this finding, and to determine a potential role for this gene in the development of schizophrenia.