Membrane-bound vascular cell adhesion molecule 1 (VCAM-1) allows the tethering and rolling of monocytes and lymphocytes as well as firm attachment and transendothelial migration of leukocytes. Soluble forms of VCAM (sVCAM-1) may serve as monitors of increased expression of membrane-bound VCAM-1 and thus may reflect progressive formation of atherosclerotic lesions. Levels of sVCAM-1 have been found to be increased among type 2 diabetic as compared with nondiabetic subjects. To study the association of plasma sVCAM-1 concentration and risk of cardiovascular and all-cause mortality among nondiabetic and diabetic subjects, we investigated an age-, sex-, and glucose-tolerance-stratified sample (n = 631) of a populationbased cohort aged 50-75 years that was followed p r o s p e c t i v e l y. Plasma levels of sVCAM-1 were determined in frozen -70°C baseline samples. After 7.4 years (mean) of follow-up, 107 (17%) subjects had died ( 4 2 of cardiovascular causes). In the entire group, increased sVCAM-1 levels were significantly associated with increased risk of cardiovascular mortality (relative risks [RRs] per 100 ng/ml sVCAM-1 increase, 1.10 [1.05-1.15] after adjustment for age, sex, and glucose tolerance status). This RR was somewhat diminished by further adjustment for the presence of hypertension and cardiovascular disease; levels of total, HDL, and LDL cholesterol and homocysteine; the presence of microalbuminuria (a putative marker of endothelial dysfunction); levels of von Willebrand factor (a marker of endothelial dysfunction) and C-reactive protein (a marker of low-grade inflammation); and estimates of glomerular filtration rate. However, the RR remained statistically significant. The RR among type 2 diabetic subjects was 1.13 (1.07-1.20) per 100 ng/ml sVCAM-1 increase after adjustment for age and sex, which was somewhat higher but not significantly different from the RR in nondiabetic subjects (P value for interaction term, 0.12). Further adjustment for other risk factors gave similar results. In conclusion, levels of sVCAM-1 are independently associated with the risk of cardiovascular mortality in type 2 diabetic subjects and therefore might be useful for identifying subjects at increased cardiovascular risk. Increased plasma sVCAM-1 levels may reflect progressive formation of atherosclerotic lesions, or sVCAM-1 itself may have bioactive properties related to cardiovascular risk. Our data, however, argue against the hypotheses of sVCAM-1 levels simply being a marker of endothelial dysfunction, of low-grade inflammation, or of an impaired renal function. D i a b e t e s 4 9 :4 8 5-491, 2000