OBJECTIVES: Bipolar disorder (BD) treatment guidelines state that antidepressants (ADs) may precipitate, or exacerbate manic or mixed episodes and generally recommend tapering or discontinuing ADs for patients with recent acute manic or mixed episodes. This study assessed the association between continued AD use and BD-related re-hospitalizations among BD patients. METHODS: Using the PharMetrics Patient-Centric Database from January 1, 2004 – June 30, 2007, we identified adults with an acute psychiatric event (“index event”), defined by hospitalizations, emergency room visits, or physician visits with a prescription for a new BD medication with a primary diagnosis of BD I mania or BD I mixed. All patients were required to be eligible for the 12 months before and after index event and have a BD-related hospitalization (any diagnosis of any BD subtype) in the pre-index period. Patients with schizophrenia at any time during the study period were excluded. Continued AD use was defined using prescription drug claims as 30+ days of available AD therapy within 120 days after the index event. Logistic regression––controlling for age, sex, geographic region, baseline comorbidities––was used to determine the association between continued AD use and post-index BD-related re-hospitalizations (any diagnosis of any BD subtype). RESULTS: A total of 2126 patients met study criteria (mean age, 44; 57% female). 433 BD patients (20.4%) were re-hospitalized within 1 year. Patients receiving continued AD therapy (31.5%) were significantly more likely to be re-hospitalized (OR, 1.387; 95% CI, 1.102, 1.748) than those without continued use. Other predictors of increased risk of re-hospitalization included being female and baseline diagnoses of comorbid substance abuse and eating disorders. CONCLUSIONS: Continued AD use in manic or mixed BD may be inappropriate. In this study, AD use for 30+ days within 120 days after an acute episode was associated with greater risk of BD-related re-hospitalization.