OBJECTIVES: The association of antihypertensive medication adherence with occurrence of complications and all-cause medical costs was examined using data from the Health Insurance Review and Assessment Service(HIRA). METHODS: This study was a retrospective analysis of HIRA administrative claims data from January 2005 to December 2009. All subjects initiating therapy with oral antihypertensive drugs were identified in January through December 2006. The index date was the date of the first hypertension-related prescription. Patients were observed for 1 year before and 3 years after therapy initiation. Outcomes included occurrence of complications and all-cause medical costs; occurrence of complications was measured starting 2 years after therapy initiation through follow-up and all-cause medical costs were measured all durations of follow-up. Adherence was measured with a medication possession ratio (MPR) and categorized <19%, 20-39%, 40-59%, 60-79%, and ≥80%. Using Cox survival regression, Occurrence of complications was modeled at levels of MPR. All-cause medical costs were modeled with generalized linear model regression. RESULTS: Average MPR in the total of 290,602 hypertension patients was 82.8%(±21.2). In Cox survival analysis, risk of complications increased as levels of MPR decreased. Patients with the lowest MPR level versus optimal adherence had 3 times higher risk of complications (hazard ratio=3.00, 95% CI=2.69-3.35). Under the adjustment with covariates, annual all-cause medical costs per patient were 23% higher for patients with the lowest MPR level versus optimal level (exponential coefficient=1.23, 95% CI=1.20-1.25). The costs, however, were 2-7% lower for groups with MPR 20-39%, 40-59% and 60-79%, because pharmacy costs offset in/out-patient care cost saving. CONCLUSIONS: This study showed strong associations between decreased antihypertensive medication adherence and increased risk of complication occurrence. And the lowest adherence was associated with higher all-cause medical costs. Further research is needed in longer than 3 year follow-ups to assess the economic impact of adherence.