Background/Aims: Medication non-adherence is a potentially modifiable determinant of cardiovascular disparities. Yet, little is known when and how to intervene, particularly when access barriers have been reduced or eliminated. The purpose of this study was to examine adherence among patients newly treated for hypertension to identify points of divergence and opportunities for early intervention. Methods: Our setting was Kaiser Permanente Northern California (KPNC), an integrated health care delivery system. In this retrospective cohort study, we included adults (=18 years) with hypertension who were new users of antihypertensive therapy in 2008. Our outcome measure was early nonpersistence, defined as failing to refill within 90 days of running out of the first prescription. We used multivariate logistic regression analysis to explore the relationships between race/ethnicity and early non-persistence, before and after adjustment for socio demographic (age, gender, income, copay requirements, BMI, smoking status) and clinical factors (baseline systolic blood pressure, physical comorbidity, psychiatric comorbidity and the number of physician visits). Results: We identified 44,167 patients newly treated with antihypertensive therapy. Racial and ethnic subgroups (white 37%, black 6.9%, Asian 8.8%, Hispanic 10.1%, other/unknown 37.2%) were similar with respect to demographic and clinical characteristics, including baseline systolic blood pressure. Across all subgroups, most (>95%) patients picked up their first prescription for antihypertensive therapy. However, more than one-third exhibited early non-persistence, which was highest among black (42.2%) and Hispanic (41.1%) patients. Black (Odds Ratio: 1.58;95% CI: 1.46-1.73), Hispanic (1.48;1.37-1.59), and Asian race (1.36 (1.26-1.47)] were independently associated with higher odds of early non-persistence with antihypertensive therapy in models adjusting for age and gender and these differences were not attenuated by the inclusion of a host of socio-demographic and clinical factors. Conclusions: In this setting, racial and ethnic differences in antihypertensive treatment adherence were present at an early stage of treatment and robust to socio-demographic and clinical adjustment, suggesting that non-clinical factors (e.g., illness beliefs, concerns about medicines) may be important primary drivers of disparities in early non-adherence. Identifying individual level barriers at the time of initiation of therapy may be an important goal of interventions to reduce disparities in adherence.