Microalbuminuria is a strong indicator of the risk of future cardiovascular disease and renal dysfunction. Slightly increased levels of homocysteine, an independent risk factor for atherothrombotic disease, have recently been found to be associated with the presence of (micro)albuminuria. However, it is unknown whether increased homocysteine levels precede the occurrence of (micro)albuminuria. Normoalbuminuric subjects (n=316, 66 with non-insulin-dependent diabetes mellitus [NIDDM]) of an age-stratified, sex-stratified, and glucose tolerance-stratified sample of a population-based cohort study were investigated at baseline and after a mean follow-up duration of 6.1 years. Development of (micro)albuminuria was defined as a mean albumin-to-creatinine ratio >2.0 mg/mmol at the follow-up examination. The cumulative incidence of (micro)albuminuria was 14. 0% (9.7 % to 18.3%) among nondiabetic subjects and 22.7% (12.9% to 32.5%) among NIDDM patients. Age-adjusted, sex-adjusted, and glucose tolerance status-adjusted logistic regression analyses showed development of (micro)albuminuria to be significantly associated with baseline homocysteine levels >19.0 micromol/L compared with homocysteine levels <9.1 micromol/L (odds ratio [OR] 5.1, 95% CI 1.1 to 23.0). For homocysteine levels of 9.1 to 14.0 micromol/L and 14.1 to 19.0 micromol/L, the values were OR 1.2 (95% CI 0.5 to 3.0) and OR 1.8 (95% CI 0.6 to 5.3), respectively. Additional adjustment for baseline insulin resistance, blood pressure, body mass index, presence of cardiovascular disease and retinopathy, current smoking, or estimates of glomerular filtration rate did not materially affect the results. Substituting homocysteine levels as a continuous variable for categories of homocysteine levels showed that a 5-micromol/L increase of the homocysteine level was associated with an increased risk of developing (micro)albuminuria (OR 1.38, 95% CI 0.97 to 1.95). Analyses performed in nondiabetic and diabetic subjects separately gave similar results among nondiabetic subjects. Among diabetic subjects, the association between homocysteine level and (micro)albuminuria could not be estimated, because there was an insufficient number of diabetic subjects with high homocysteine levels. Hyperhomocysteinemia is an independent determinant of the development of (micro)albuminuria among nondiabetic subjects, even after adjustment for estimates of glomerular filtration rate. We could neither confirm nor reject an association between homocysteine levels and the development of (micro)albuminuria among NIDDM subjects. These data suggest that homocysteine may play a pathophysiological role in the development of (micro)albuminuria.