The family, twin, and adoption studies of the DSM-III-R personality di
sorders are reviewed. Most data concern the schizotypal, borderline, a
nd antisocial personality disorders, where increased family risk has b
een reported. Only antisocial personality disorder has so far shown co
nvincing data for a genetic etiological factor. Most family studies in
the field have used inadequate methodology, and methodological improv
ement is a major issue of such studies in the future.