Hypertension is a major public health problem with serious medical and financial consequences. Barriers to successful conventional pharmacological treatment include side effects, out-of-pocket expenses, patient noncompliance and insufficient dosages. Acupuncture has been studied as an alternative therapy for controlling blood pressure (BP) but previous studies have serious methodological limitations. This paper describes the design of the Stop Hypertension with the Acupuncture Research Program (SHARP) trial, a pilot randomized clinical trial designed to gather preliminary data regarding the efficacy of traditional Chinese medicine (TCM)-based acupuncture for control of essential hypertension. The design of the SHARP trial balanced rigorous clinical trial methodology with principles of TCM. Eligible participants had systolic BP (SBP) 140–179 mm Hg and diastolic BP (DBP) 90–109 mm Hg in the absence of antihypertensive therapy. Following screening, participants were randomized to one of three groups: individualized, standardized or control acupuncture. Treatments were designed according to principles of TCM; nonspecific effects associated with the interventions were standardized across the randomized groups. For individualized acupuncture, points were tailored to each participant. Standardized acupuncture used a prespecified set of points. The invasive sham control acupuncture regimen was designed to be non-active. Each participant received a “prescription” for individualized acupuncture from an acupuncturist who was masked to treatment assignment, and was subsequently treated by an independent acupuncturist. Patients and those assessing BP were masked to treatment group. Acupuncture was delivered twice a week for 6 weeks. Follow-up visits were every 2 weeks to week 10 and then at months 4, 6, 9 and 12. The primary endpoint will be change in SBP from baseline to 10 weeks. DBP, BP trajectories over the 12-month follow-up and antihypertensive medication requirements will also be examined. Initial contact was documented for 1442 prospective participants from March 2001 to April 2002; 424 provided informed consent and 192 were ultimately randomized.