Objective: To compare post-treatment costs across alternative antipsychotics in the treatment of bipolar disorder (BD). Methods: Data from a commercial health plan from July 1, 2003 to June 30, 2006 were used to identify non-institutionalized patients with bipolar disorder (ICD-9 296.4-296.8) but no history of schizophrenia (ICD-9 295.xx). Patients initiating treatment using a typical antipsychotic (TAP), atypical antipsychotic (AAP: aripiprazole, olanzapine, quetiapine, risperidone or ziprasidone), mood stabilizer or antidepressant were included. Episodes were divided into three categories: restarting treatment after a break in drug therapy > 15 days with the drug used in the previous episode, switching therapy with or without a break in treatment, and augmentation therapy. First observed episodes were excluded from the analysis due to uncertainty concerning the patient's prior treatment history. A total of 106,447 episodes were included in the analyses using ordinary least squares (OLS) regression models of post-treatment cost adjusting for age, gender, geographic region, drug use history, prior medical care use, bipolar disorder diagnosis and co-morbid medical conditions. Results: Average total post-treatment cost measured across all episode types ranged from $17,837 (olanzapine) to $22,292 (ziprasidone). OLS results found augmentation episodes to be significantly more costly than restart episodes (+$6836, p<0.0001) or switching episodes (+$4109, p<0.0001). AAPs were found to be more costly relative to TAP in patients restarting therapy and these estimates were significant for quetiapine (+$3126, p<0.01) and ziprasidone (+$4811, p<0.05). Patients initiating augmentation episodes with an AAP were also consistently more costly relative to TAP, again significantly so for quetiapine (+$2534, p<0.05) and ziprasidone (+$2846, p<0.05). However, most AAPs achieved significantly lower total costs relative to TAP for switching episodes ranging from -$1,817 for ziprasidone (p>0.05) to -$7632 (p<0.0001) for olanzapine. Conclusion: In a commercially-insured population, AAPs are only associated with lower total post-treatment costs in patients with bipolar disorder who switch therapies.