OBJECTIVES: Recent studies have demonstrated that up to 30% of patients never fill an initial prescription for antihypertensive medication. This high rate of primary nonadherence has direct implications for the outcomes of clinical care. Dramatic improvements in the availability of administrative pharmacy claims in the electronic health record (EHR) may promote the identification of nonadherent patients. The objective of this study was to characterize pharmacy fill patterns for a newly prescribed antihypertensive. METHODS: We conducted a retrospective cross-sectional study on patients prescribed a new antihypertensive within a large primary care practice network from 2011-2012. Patients were included if they had a diagnosis of hypertension or elevated blood pressure at the time of the new prescription and at least one practice visit within the previous 18 months.  We excluded patients who did not have an imported pharmacy fill history on or after the antihypertensive index date.  Pharmacy fill claims were pulled for the subsequent 90 days after the index date. All fills of the antihypertensive were then linked to the antihypertensive as found in the EHR. RESULTS: A total of 131 patients and 133 medications were studied. Average age (SD) was 55.4 (14.9), 74.8% were female, 26.7% were black and 32.8% were white. Most patients were prescribed one antihypertensive; 2 patients were prescribed two. Of the medications studied, 112 medications were linked to a fill from the pharmacy claims. Of the linked medications, 73.2% were filled on the index date. Average number of days (SD) to first fill was 4.6 (12.1); 94.6% of linked fills were filled within the first 30 days. CONCLUSIONS: Nearly a quarter of patients do not fill newly prescribed antihypertensives within 90 days of initial prescription. Our preliminary data suggest that the real-time availability of pharmacy claims data could be useful to identify nonadherent patients.