Objective, fast and easy methods have not been established in the examination of urine sediment to differentiate between glomerular and nonglomerular hematuria. In this study, we used a newly developed automated urinary flow cytometer that can clearly recognize red blood cells (RBC), white blood cells, epithelial cells, bacteria and crystals by their size and fluorescence intensity without sedimentation. 98 urine samples from 31 glomerular and 67 nonglomerular lesions were analyzed by the device, and the criteria to determine the origin of hematuria were established based on the results. Additional 108 cases were tested to evaluate the validity of these criteria. According to the analysis of histograms of urinary RBC size distribution, cases in whom ≥80% of all RBC have forward scatter (FSC) intensities ≦126 and <80% of all RBC have FSC intensities ≥84 were regarded as representing the glomerular type. Cases in whom ≥80% of all RBC have intensities FCS ≥84 and <80% all RBC have FSC intensities ≦126 were regarded as representing the nonglomerular type. Cases in whom <80% of all RBC have FSC intensities ≦126 and <80% of all RBC have FCS intensities ≥84 were regarded as the mixed type. Cases in whom ≥80% of all RBC have FSC intensities ≦126 and ≥80% of all RBC have FSC intensities ≥84 were regarded as the nonglomerular type. The sensitivity for glomerular RBC in the first 98 cases was 90.3% and the specificity 92.5%, and in the second 108 cases the values were 100 and 86.6%, respectively. The automated urinary flow cytometer is useful as a means for routine differential diagnosis of hematuria, and at least it is promising as the screening test for differentiation between glomerular and nonglomerular hematuria, because it can examine numerous samples within a short time and does not necessitate any special skill or knowledge.