A recent case-control study suggested that modest enlargements of a CAG repeat in the hKCa3 potassium channel may be associated with bipolar disorder. We tried to replicate this result in a UK Caucasian sample of 203 DSM-IV bipolar I disorder patients and 206 controls group-matched for age and sex. Using the same model of analysis as the earlier study, bipolar probands did not have a higher frequency of alleles with greater than 19 repeats than controls (chi2 = 1.44, 1 df, P = 0.23). Similarly, comparison of the distributions of repeat sizes between probands and controls did not approach statistical significance (Mann-Whitney U test, P = 0.35). We conclude that our data provide no support for the hypothesis that variation at the hKCa3 gene contributes to susceptibility to bipolar disorder.