This study tests the hypothesis that spironolactone influences plasminogen activator inhibitor-1 (PAI-1) concentrations through mineralocorticoid receptor antagonism rather than through changes in potassium. Effects of spironolactone (50 mg per day) and triamterene (50 mg per day) on fibrinolytic balance were compared in 18 normotensive and 20 hypertensive subjects pretreated with hydrochlorothiazide (HCTZ; 12.5 mg per day). Blood pressure and serum potassium were similar in spironolactone and triamterene treatment groups. The effect of the 2 drugs on the renin-angiotensin-aldosterone system was also similar. In contrast, spironolactone and triamterene exerted opposing effects on PAI-1 antigen (P=0.006 for drug effect). In normotensive subjects, triamterene (from 10.1+/-7.8 to 16.9+/-9.9 ng/mL at 9 am, P=0.019; from 7.6+/-5.4 to 11.5+/-7.3 ng/mL at 11 am, P=0.027; from 9.3+/-7.7 to 13.7+/-8.5 ng/mL for average of all time points, P=0.054) but not spironolactone significantly increased PAI-1 antigen. In hypertensive subjects, spironolactone significantly decreased PAI-1 antigen (from 22.0+/-23.4 to 16.7+/-19.0 ng/mL at 10 am, P=0.041; from 17.5+/-21.7 to 12.7+/-16.8 ng/mL at 11 am, P=0.043; from 20.3+/-22.6 to 16.6+/-19.7 ng/mL for average of all time points, P=0.014), whereas there was no effect of triamterene. Only spironolactone significantly decreased the molar ratio of PAI-1 to tissue-type plasminogen activator (t-PA) in hypertensive subjects. By regression analysis, predictors of mean PAI-1 response were spironolactone versus triamterene (P=0.014), hypertension (P=0.002), and PAI-1 response to HCTZ (P=0.019), with a trend for aldosterone (P=0.061). Mineralocorticoid receptor antagonism prevents the effect of activation of the renin-angiotensin-aldosterone system on PAI-1 antigen in normotensive subjects and improves fibrinolytic balance in hypertensive subjects through a potassium-independent mechanism.