OBJECTIVES: This study compares the risks of side effects [SE] associated with a broad of range of atypical and typical antipsychotics used to treatment patients with bipolar disorder.  METHODS: Medical and pharmacy claims data from Humana from January 2007 to June 2013 were used to define episodes of antipsychotic drug therapy for patients with bipolar disorder.  Episodes were screened for a minimum of 180 days of pre-episode and 360 days of post-episode data, and for the existence of study SEs prior to the episode index date.  Study samples ranged from N=147,658 for rhabdomyolysis [RM] to N=150,084 acute urinary retention [AUR].  Other SE for analysis included acute kidney injury [AKI], hypotension, pneumonia (PNA), acute coronary syndrome (ACS) or ischemic stroke/cerebral infarction, ventricular (tachy) arrhythmia (VCA).  Logistic regression models were used to estimate the impact of alternative antipsychotics on the risk of an emergent SE in the year following the initiation of treatment.  The logistic models controlled for patient demographics, treatment history, concomitant use of antidepressants, mood stabilizers and anticonvulsants, diagnostic and drug use profiles in the prior 6 months.  Haloperidol was used as the comparison drug. RESULTS: Aripiprazole, ziprasidone and risperidone did not differentiate relative to haliperidol.  Only fluphenazine increases the risk of cardiovascular events relative to heliperidol.  All other SE impacts related to specific antipsychotics were related to kidney disorders or pneumonia.  Chlorpromazine increased the risk of acute kidney injury, hypotension, rhabdomyolysis and pneumonia. Clozapine increased the risk of pneumonia.    Olanzapine increased the risk of AKI and pneumonia but significantly decreased the risk of AUR.  Paliparidone also increased the risk of PNA and decreased the risk of AUR.  Quetiapine increased the risk of AKI while perphenazine decreased the risk of hypotension and thiothixene decreased the risk of AUR. CONCLUSIONS: The risk of side effects varies widely across antipsychotic medications.