Evidence about the genetic determinants for borderline conditions is reviewed. The research data are too limited and the diagnostic practices followed in existing studies are too varied to allow firm conclusions to be drawn. Thus, the need for new studies starting with well-defined samples of borderline patients is clear. Previous work implicates genetic factors in the etiology of at least some borderlines, but it is unclear that borderlines by any definition will, as a group, have uniformly strong and specific genetic determinants. With further research, partially overlapping subgroups might be defined on the basis of careful examination of borderline patients' clinical characteristics and family histories. Such characterization could have potential clinical value since there may be subgroups of borderlines who respond differently to various psychopharmacologic treatments (Klein 1975) or to exploratory versus ego supportive psychotherapy (Stone 1977). Characterizations based on genetic considerations may have their limitations, since the genetics of these disorders probably involve complex interactions of a variety of factors that may be more or less specifically related to other major diagnostic groups.