Conclusions: Ziprasidone, alone or added to antidepressants, was not more effective than placebo in this population. A false-negative finding due to the crossover design is suggested by statistical carryover effects. Alternatively, this definition of bipolar spectrum illness may have been too nonspecific to show neuroleptic benefit, unlike other definitions, like "mixed depression." Also, this study did not test potential neuroleptic efficacy without the potentially mood-destabilizing effects of antidepressants.