OBJECTIVES: To compare adherence rates and health care utilization for clozapine versus other second-generation antipsychotics used as monotherapy in the treatment of schizophrenia.   METHODS: The study was an observational, retrospective analysis of medical and pharmacy claims from a national insurance organization with approximately 14 million enrollees, conducted using the i3 InvisionTM Data Mart database for dates of service from January 2006 through December 2009. Study patients had >1 pharmacy claim for a second-generation antipsychotic (study drugs: apiprazole, clozapine, olanzapine, paliperidone, quetiapine, risperidone, or ziprasidone) with no pharmacy claims for any study drug during a 6-month preperiod prior to the initial (index) claim. All patients met the following criteria: no pharmacy claims for any second-generation antipsychotic other than the index medication during the 90 days following the index claim; aged 18 to 64 years; continuously enrolled from the preperiod through 12 months after the index date (postperiod); and >1 medical claim with a diagnosis of schizophrenic disorder (ICD-9-CM=295.xx) in either period. Logistic regression analyses modeled binomial measures of postperiod medication possession ratio (MPR) for the index drug (>80%=adherent) and postperiod hospitalizations and emergency room (ER) use in 3 categories: all-cause, mental disorder-related (ICD-9-CM=290.xx-319.xx), and schizophrenia related (ICD-9-CM=295.xx), controlling for patient demographics, preperiod hospitalization and ER use, and preperiod comorbidities.   RESULTS:  Mean MPRs for the sample (N=19,804) were 0.78 for clozapine, 0.62 for quetiapine, and 0.42-0.49 for the remaining study drugs. In logistic regression analyses, the odds of MPR >80% were lower for all study drugs compared with clozapine (odds ratio[OR]=0.603, 95% confidence interval=0.483-0.754 for quetiapine and ORs=0.306 to 0.390 for remaining drugs).  CONCLUSIONS: Compared with other second-generation antipsychotics, treatment with clozapine is associated with better adherence and equal or less all-cause and disease-specific hospital care utilization. These findings should be interpreted in light of the benefits and risks of all antipsychotic drugs.