The increased cardiovascular risk associated with rofecoxib may relate to differential effects on blood pressure (BP) in comparison with other cyclooxygenase-2 (COX-2) selective inhibitors. Rofecoxib is uniquely metabolized by cytosol reductase and may compete with aldosterone for metabolism. We hypothesized that the effect of rofecoxib on BP may be due to an increase in aldosterone levels. Prospective, randomized, cross-over study. Tertiary institution. Eleven patients (all female, age 65 +/- 11 years) with osteoarthritis and hypertension. Patients received rofecoxib 50 mg once/day or celecoxib 400 mg once/day for 8 weeks, followed by cross-over after a 2-week wash-out period. Office BP, heart rate (HR), plasma aldosterone and aldosterone related-markers, markers of sympathetic and parasympathetic activity, catecholamines and heart rate variability (HRV) were measured. Rofecoxib caused an increase in BP compared to celecoxib; a change in recumbent systolic BP +/- SD (6.1 +/- 11.0 versus -7.0 +/- 12.5 mmHg, P < 0.05) and recumbent diastolic BP +/- SD (3.6 +/- 7.6 versus -1.7 +/- 5.7 mmHg). There was no difference in median change in plasma aldosterone levels between rofecoxib and celecoxib (-31 +/- 171.5 versus 43 +/- 147 pg/ml, P = 0.173). The other parameters measured were not different between the two agents. Rofecoxib therapy results in a significant increase in systolic BP compared to celecoxib. As plasma aldosterone levels were not different between the two drugs, the differential effect of rofecoxib on BP compared to celecoxib appears to be unlikely due to competition by rofecoxib for aldosterone metabolism.