Although published literature shows a clear relationship between poor patient compliance with antihypertensive medication and failure to achieve blood pressure control, the association between treatment choices and compliance is less clear. Patient behavior in randomized controlled trials differs from that in actual practice; studies of this difference are therefore required to consider real-world evidence. Studies of actual practice have used a variety of measures of patient behavior, which makes comparisons difficult. In addition, they often have mingled patients receiving antihypertensive treatment for the first time and established hypertensive patients--groups that generally have quite different levels of compliance. This article briefly reviews the literature concerning therapeutic choices and patient compliance with hypertensive medications, including a recent study of persistence using treatment data from the provincial health plan of Saskatchewan, Canada. The Saskatchewan analysis was a rigorous examination of actual practice treatment for hypertension, which demonstrated that angiotensin-converting enzyme inhibitors were associated with the highest 1- and 4.5-year persistence rates and the lowest discontinuation rates among calcium antagonists, beta-blockers and diuretics. These findings may hold important implications for physicians when choosing an initial antihypertensive therapy, and point to the potential importance of the excellent safety and tolerability profiles of new antihypertensive therapies.