From family, adoption, and twin studies it is clear that genetic factors play an important role in the etiology of bipolar disorder (McGuffin and Katz: The Biology of Depression, Gaskell, London, 1986). Recently Yoneda et al. (Am J Hum Genet 51:1172-1173, 1992) reported an association between an allele (A4) of a VNTR marker (DXYS20) for the pseudoautosomal region and bipolar disorder in a Japanese population. In order to test for this association in a Caucasian population, we have typed a sample of 52 subjects with bipolar disorder and 61 normal controls. The bipolar subjects are probands of multiple incidence families. The normal controls are an epidemiologically ascertained sample of middle-aged, unrelated individuals. The two groups were matched for sex and ethnic background. There were no significant differences in the allele or genotype frequencies of DXYS20 between the two groups. In particular, there was no significant difference in the frequency of the A4 allele in normal controls and bipolar patients (0.377 vs. 0.317, respectively). The prevalence of the A4 allele in bipolar patients and normal controls was 0.567 and 0.622, respectively. We were not able to replicate the results of the Yoneda et al. (1992) study.