Beginning in 1974, the Japanese Ministry of Health Welfare directed the screening of schoolchildren for proteinuria. We studied their procedure and methods in 6197 school children and also evaluated a new urine dipstick that measures albumin concentrations down to about 10 mg/l and creatinine down to about 300 mg/l. We used specimens from adult in- and outpatients to test the accuracy of the dipsticks. Based on the quantitative results, we set as cutoffs < 150 mg/l for protein and < 30 mg/l for albumin as the concentrations representing "low risk." The quantitative values were assumed to be correct, and the dipstick results were judged accordingly, i.e., a dipstick protein of > or = "150" mg/l or an albumin of I "30" mg/l indicated increased risk of developing or having a genitourinary disorder. The sensitivity/specificity of the protein dipstick was 95.1%/95.5%, and the same for the albumin dipstick was 83.8%/93.8%. The cut-off for the albumin dipsticks probably should be set somewhat lower to reduce the number of false negatives and increase the sensitivity of the dipstick. When we compared the quantitative albumin to the protein dipsticks with the above cut-offs, we found the sensitivity/specificity to be 79.3%/94.4%, i.e., much like the albumin dipstick results. The many reports on the association of albuminuria and risk of renal disease recommend that screening should be done for albumin rather than protein. Based on the data from the school children, we estimate that a dipstick albumin of "30" mg/l is borderline increased risk, and that a protein dipstick of "150" mg/l is the same. If we call the dipstick "10" mg/l albumin, "30" mg/l albumin and the "150" mg/l protein results "low risk," then we estimate the prevalence of albuminuria in the school children to be about 2.1% and proteinuria to be about 4.3%. Children with these values should have a quantitative test for albumin and protein. We also tested a dipstick for creatinine and found increasing values with increasing age in both genders; the older boys had significantly higher creatinine values than the older girls and younger boys. For the albumin/creatinine ratio, we found 6028 children with a ratio of < 30 mg/g indicating low risk and 159 children with a ratio of > or = 30 mg/g indicating increased risk. The ratio may be more useful owing to the likely reduction of the number of false negatives and false positives.