Over the last 20 years several pedigrees have been reported which are suggestive of linkage between susceptibility to bipolar disorder and markers on chromosome Xq28. Other workers have failed to replicate these reports and the methodology of the positive reports has been criticised. Recently there have been several reports of an association between fragile X (FRA(X)) and affective disorder within families and in unrelated individuals compared with controls. Such reports could be consistent with the Xq28 marker reports because FRA(X) maps to Xq27.3. We report a study at the FRA(X) CGG repeat locus in 79 unrelated Caucasian bipolar probands without fragile X syndrome and 77 unrelated controls. We found no evidence that variation at this locus confers susceptibility to bipolar disorder.