Research has suggested the existence of a pathological dissociative taxon. However, relatively little is known about this taxon. This study examined the 2-month retest stability of this taxon--together with other measures of dissociation and the Big Five--in a sample of 465 undergraduates. Contrary to expectation, taxon scores were only modestly stable and were substantially less stable than the other measures, including continuous indicators of dissociation. Furthermore, most individuals who were identified as taxon members at one assessment were classified as nonmembers at the other. These results challenge the existence of a pathological dissociative taxon. More generally, these data demonstrate that statistically identified taxa need to be explicated through the normal process of construct validation.