High lod scores were obtained in several X-linkage studies of bipolar illness under the assumption that a subgroup of manic depression follows an X-linked dominant mode of inheritance. We have previously shown that the segregation patterns do not substantiate this assumption. We now statistically address the lack of evidence for X-linked inheritance by sex-dependently analyzing segregation ratios and clinical data presented in eight positive X-linkage studies. Accordingly, affected males have significantly fewer offspring, a lower mean age and a higher bipolar to unipolar ratio than affected females. There are two possible explanations for these findings: either the X-linked subgroup of bipolar illness has unique features that have not been accounted for clinically, or (more probably) the pedigree structures could also (and might be more likely to) result from ascertaining kindreds following autosomal or multifactorial inheritance only, with exclusion of kindreds encompassing male to male transmission.