The main conclusions of this study on the familial links between personality patterns and affective disorders are: (1) The personality features with the greatest degree of symptomatic overlap with unipolar depression were more common among the first-degree relatives of probands with this diagnosis: thus dysthymic temperament and neuroticism are enhanced in this group of relatives compared to controls. Likewise personality features with a high degree of symptomatic overlap with bipolar affective disorder were more common among the first-degree relatives of probands with this diagnosis. Thus levels of dysthymic and cyclothymic temperament were elevated in this group of relatives compared to controls, whereas a familial link between neuroticism and bipolar disorder was not observed. (2) In addition, personality traits with only limited similarity with the syndromes of depression and mania were also found to be linked with affective disorders (obsessive-compulsive PD to bipolar disorders, rigidity to both subtypes of affective disorders). These associations have different implications regarding the association between personality/temperament conditions and disorders depending on the degree of overlap between symptoms and the personality trait. Those with substantial overlap may indicate the presence of a unitary disease process in which the enhancement of subthreshold affective traits may result in aggravation of behavior characteristics fulfilling the criteria of fullblown episodes of the disease. This relationship was particularly stressed by Kraepelin, Kretschmer, Clayton et al. and Akiskal. In this view personality and temperament patterns represent minor variants of the associate acute disorders. On the other hand, the relationship between dissimilar conditions may indicate the presence of a risk of underlying vulnerability factors which led to affective disorder only in the presence of additional risk factors. An example of the latter relationship in the present study is the aggregation of obsessive-compulsive and anancastic traits in families with affective disorders. Tellenbach focussed on this particular constellation. Epidemiological and family studies including these personality traits are too rare to fully appreciate the relevance of this particular relationship. Future prospective studies and family genetic studies which investigate the relationship between temperament, personality traits and disorders and affective syndromes are clearly indicated by the results presented herein.