To explore the prevalence of hematuria or proteinuria in school children in Shanghai and to evaluate the screening methods, we conducted urine screening in more than 40,000 school children between 2003 and 2005. Children were tested with dipsticks read manually (method A) or dipsticks read by machines (method B) combined with a sulfosalicylic acid test or microscopy. Some children were tested once, and others who had abnormal results in the first screening were tested again 2 weeks later. The prevalence of urine abnormalities in the first screening was more than 5.00% and of the second screening about 1.00%. Either method B or testing two urine samples for each child had higher specificity. As to the direct cost, that of screening twice with method A was lower than just screening once with method B. So using method A to screen twice for each child was not only convenient and economical, but also could reduce the false positive rate effectively. More than 10 months of follow-up diagnosed two cases of IgA nephropathy. Asymptomatic chronic renal diseases in school children could be detected through school urine screening. For Shanghai, China, screening twice using method A might be the best choice.