Background. Mild renal insufficiency has recently been recognized as an important risk factor for cardiovascular disease (CVD). The mechanisms underlying this association are incompletely understood. Increased left ventricular mass (LVM) is an independent risk factor for CVD, which is particularly common in end-stage renal disease ( ESRD) and which has been shown to be associated with mild renal insufficiency. Increased arterial stiffness has also been shown to be an independent risk factor for CVD in ESRD and has also been associated with mild renal insufficiency. We hypothesized that the association between mild renal insufficiency and increased LVM could be mediated through increased arterial stiffness, and that this may be one of the pathways linking mild renal insufficiency to CVD. We therefore investigated, in a cross-sectional population-based study, the influence of increased arterial stiffness on the association between renal function and LVM. Methods. The study population consisted of 742 elderly individuals ( 373 men and 369 women). Renal function was estimated by the serum creatinine level in mu mol/L; by the Cockcroft-Gault formula in mL/min and by the Modification of Diet in Renal Disease (MDRD) formula. LVM was obtained by echocardiography. Results. The mean estimates of renal function in men and women were, respectively, 103.7 (SD 17.0) and 86.8 (SD 11.2) mu mol/L for the serum creatinine level; 63.4 ( SD 12.9)and 61.4 ( SD 11.0) mL/min/ 1.73m(2) for the Cockcroft-Gault formula; and 59.7 ( SD 10.8) and 60.9 ( SD 10.5) mL/min per 1.73 m(2) for the MDRD formula. LVM was 93.1 ( SD 26.4) g/m(2) in men and 86.7 ( SD 22.3) g/m(2) in women. In men, impaired renal function, as estimated by the Cockcroft-Gault and the MDRD formula, was significantly associated with greater LVM after adjustment for age, glucose tolerance, hypertension, and prior CVD [ regression coefficient beta ( 95% CI), 1.28 ( 0.22 to 2.33) g/m(2) and 1.63 (0.41 to 2.86) g/m(2) per 5 mL/min/1.73m(2) decrease, respectively]. However, the association between impaired renal function and increased LVM was not statistically significant after adjustment for arterial stiffness estimates [regression coefficient b ( 95% CI), 0.02 (-1.60 to 1.64) g/m(2) and 0.54 (-1.25 to 2.33) g/m(2) per 5 mL/ min/1.73 m(2) decrease, respectively]. In women, impaired renal function was not significantly associated with greater LVM. Conclusion. Our study shows that in a general elderly population, even mild impairment of renal function is associated with adverse changes in left ventricular structure. In men, but not in women, this leads to greater LVM, a process that may be related to increases in arterial stiffness. Importantly, these novel findings suggest that such changes occur early in the process of renal functional deterioration, which may explain, in part, the increase in cardiovascular risk in men with mildly impaired renal function.