INTRODUCTION: Simple logic would suggest that there should be some endophenotype for bipolar disorder. Possible endophenotypes could include specific variations in personality. Bagby and Ryder summarized the work up to that point by noting that the related personality traits of high neuroticism and harm avoidance seem to be associated with bipolar disorder as well as with unipolar depression, whereas higher novelty seeking may be associated only with bipolar patients. As these parameters are all very sensitive to the affective state, it is critical to examine the literature that pertains specifically to euthymic patients in order to evaluate the extent to which this signifies underlying personality (trait), and not primarily clinical status (state). Several important studies have been published since the Bagby and Ryder paper, which we review here. We restrict our current review to empirical studies which employed both adequate samples of euthymic (to minimize the state/trait dilemma) bipolar patients as well as healthy comparison subjects. This paper is restricted to frequently used explicit measures of personality - that is, self-report questionnaires: the Temperament and Character Inventory (TCI) based on Cloninger's psychobiological theory of temperament and character, the Revised NEO Personality Inventory based on the five-factor model of Costa & McCrae, and the Barratt Impulsiveness Scale (BIS-11) (23). No single dimension of 'personality' reviewed would qualify as a psychological marker for a bipolar disorder. Earlier findings as reviewed by Bagby and Ryder, of higher novelty seeking, were not replicated in these studies. Of the personality traits considered, the most promising candidate for marker or endophenotype would seem to be "impulsivity" as measured by the BIS-II.