Background: The aims of this study were to identify the dominant affective temperamental characteristics of patients with bipolar disorder (BP) and their clinically well first-degree relatives and to compare the prevalence rates of these temperaments with those in healthy control subjects. Methods: One hundred bipolar I probands and their 219 unaffected first-degree relatives were enrolled in the study. The control group consisted of healthy subjects without any personal or family history of bipolar disorder, matched with the age and gender of the probands and first-degree relatives. To identify the dominant affective temperaments, the Turkish version of TEMPS-A scale was used. Results: At least one dominant temperament was found in 26% of the proband group, in 21.9% of the relative group, and 6.0% and 10.0% of the control groups, respectively. The most noteworthy finding, was that both the probands and their relatives had significantly higher frequency of hyperthymic temperament than the controls. Limitations: Temperament had not been assessed premorbidly in the probands with bipolar disorder. Conclusions: The study supports the familial, possibly genetic, basis for the hyperthymic temperament in the genesis of bipolar I dosorder. That the cyclothymic temperament was not similarly represented, may be due to the higher specificity of the cyclothymic temperament to the bipolar II sybtype (which we did not study). More research is needed on the relevance of cyclothymic and other temperaments to the genetics of bipolar disorders selected by rigorous subtyping along the clinical spectrum of bipolarity. (c) 2004 Elsevier B.V. All rights reserved.