To identify the biological covariates of microalbuminuria (albuminuria ≥15 &mgr;g/min) in nondiabetic subjects, brachial blood pressure, echocardiographic left ventricular mass, and other cardiovascular and metabolic parameters were evaluated in 211 untreated males (38 normal controls, 109 uncomplicated stage 1 to 3 essential hypertensives, and 64 patients with clinically stable atherosclerotic peripheral vascular disease either with [n=44] or without [n=20] essential hypertension) with normal cardiac and renal function. Compared with normoalbuminuric subjects, microalbuminuric subjects (n=67) were characterized by higher systolic blood pressure, comparable diastolic blood pressure, and, therefore, wider pulse pressure. Greater prevalence of hypertension, peripheral vascular disease, left ventricular hypertrophy, and reduced HDL cholesterol values further distinguished microalbuminuric from normoalbuminuric subjects in univariate comparisons. The risk of microalbuminuria increased by ascending pulse pressure quintiles in age-corrected logistic regression models, in which pulse pressure was more predictive than systolic pressure and was independent of mean pressure. When microalbuminuric status was regressed against a series of dichotomous (vascular and active smoker status) and continuous (age, pulse and mean pressure, left ventricular mass index, and HDL and LDL cholesterol) variables, only pulse pressure, left ventricular mass index, and smoking status were independent predictors. The association of increased albuminuria with wider pulse pressure, a correlate of the pulsatile hemodynamic load and conduit vessel stiffness as well as an important cardiovascular risk factor, may explain why microalbuminuria predicts cardiovascular events in nondiabetic subjects. The independence from concomitant vascular disease also suggests that wider pulse pressure, rather than representing a simple marker for atherosclerotic disease, influences albuminuria directly.