OBJECTIVES: The purpose of this study was to 1) determine medication adherence in hemodialysis patients initiated on either calcium acetate or sevelamar, and 2) determine the association between adherence to phosphate binder therapy and age in hemodialysis patients. METHODS: Using records from incident hemodialysis patients in the Centers for Medicare and Medicaid Services End-Stage Renal Disease database linked with Medicare Part D data, we compiled information on patients treated with phosphate binders (1/2007 - 6/2009). We included patients that were treated with either calcium or sevelamar. Prescription refill records in the 6-month post-index period were assessed and two measures of adherence; proportions of days covered (PDC) and total gap were computed. A PDC of ≥ 80% was considered adherent. RESULTS: A total of 7299 patients that did not switch therapy and survived at least six months from the index phosphate binder date were included in this analysis. Compared to calcium patients, sevelamar patients had a significantly higher mean PDC, a significantly lower mean total gap, were significantly more likely to be adherent; and were significantly less likely to have gaps in medication possession during the first 6 months of pharmacotherapy (P<0.05). After adjustment for multiple factors, older patients (age >= 75 at dialysis initiation) had a significantly higher mean PDC, a significantly lower mean total gap, were significantly more likely to be adherent; and were significantly less likely to have gaps in medication possession during the follow up period than did younger patients (age <75) (P<0.05). CONCLUSIONS: Overall about one-third of the patients included in this study were adherent at 6 months of initiation of phosphate binder therapy (PDC>=80%). Patients treated with sevelamar were more adherent than those treated with calcium, and they had fewer gaps in medication possession. Older patients were more adherent than younger patients.