Albumin-creatinine ration (ACR) has been correlated to increased morbidity and mortality in diabetic as well as in nondiabetic populations. We investigated the albumin-creatinine ratio in 898 randomly selected persons aged 30-50 years, 471 women and 424 men (year 0). ACR was remeasured 1 year later in 811 (90%) of these persons (year 1). This was done in the framework of a prospective, randomized, population-based intervention trial evaluating the effect of health test and health conversations in general practice. The mean age of the study population was 39.8 (range, 30-51) years with a mean body mass index of 24.3 (15.9-44.3) kg/m2. The 50th, 95th, and 99th percentiles of ACR were 0.6, 2.0, and 5.0 mg/mmol at year 0 and 0.5, 2.0, and 5.3 mg/mmol at year 1. There was no significant difference between ACR values at year 0 and year 1. ARC values were slightly higher in women than in men, 0.6, (0.2-20.4) mg/mmol versus 0.5 (0.2-16.3) mg/mmol (p < 0.001, median and range). Albumin concentrations were slightly higher in men than in women 8.0 (2.0-196.0) mg/mmol versus 6.5 (2.0-121.0) micrograms/mL, p < 0.001. The correlation between ACR values measured year 0 versus year 1 was 0.43, p < 0.001 (Spearman rho). The agreement from year 0 to year 1 between normal ACR values (< or = 2.5 mg/mmol) and abnormal values (> 2.5 mg/mmol) were 0.33 expressed by Cohen's kappa.(ABSTRACT TRUNCATED AT 250 WORDS)