BACKGROUND Mood-incongruent psychosis during bipolar disorder has been associated with poor outcomes. However, it remains unknown whether this is secondary to persistent affective or psychotic symptoms or both. METHODS Sixty-eight patients with bipolar disorder between the ages of 16 and 45 were recruited during their first psychiatric hospitalization for mania. These patients were evaluated using structured and semi-structured clinical interview then followed longitudinally. Outcomes during the first twenty-four months of follow-up were compared between patients with mood-incongruent psychosis and those without (i.e., patients with mood-congruent psychosis or no psychosis) during the index manic episode. Specifically, ratings of the percent of weeks during follow-up with the duration of mood incongruent psychotic symptom, any psychotic symptom, affective syndromes, and scores of global outcomes were compared. RESULTS Comparing the 24-month follow-up results between the two groups, patients with mood incongruent psychotic symptoms had a lower global functional rating scale, efficacy index, while the duration of mood incongruent psychotic symptom, any psychotic symptom, and complete affective symptom showed statistically significant differences between the two groups. There were also statistically significant differences between the two groups in the duration of mood stabilizers, and antidepressants use, typical antipsychotics, and atypical antipsychotics. Partial correlation analysis reveals the scores of the global assessment of functioning scale (GAF) after 24 months showed a significant negative correlation with the length of time of incongruent psychotic symptoms. Still, the correlation was intermediate (correlation coefficients less than 0.5，r2 = -0.471, P < 0.001). CONCLUSIONS Mood-incongruent psychosis that occurs during the first manic episode appears to predict an increased likelihood of persistent psychotic symptoms during the subsequent twenty-four months. This persistence of psychosis is associated with a worse overall course of illness than patients without mood-incongruent psychosis. CONCLUSIONS These results apply to a relatively short outcome period, and the sample size is relatively small.