Evidence of linkage in families of bipolar patients has so far been identified with genetic markers on chromosome X and 11. However, replications of these data have not consistently been reported in either case, which favours the hypothesis of genetic heterogeneity. Therefore, we have tried to outline a sampling strategy for linkage replication in affective disorders. We estimated the average number of nuclear families required to replicate X or 11 linkage as a function of the degree of heterogeneity as well as the number to prove heterogeneity given that linkage exists. The results are presented and discussed.