To aid investigation into the clinical problem of hematuria, assessment of abnormalities in the shape of red cells in the urine (dysmorphism) is gaining popularity in nephrology. However, there is uncertainty in the literature regarding both the number of red blood cells (RBC) in normal urine, as well as the quantification of dysmorphism. We have shown that in normal urine (N = 27) the number of RBC is less than 2,000/ml as assessed by scanning electron microscopy of filtered urine specimens from normal volunteers without known renal disease, which compared to less than 1,000/ml by centrifugation and phase contrast microscopy of the same specimen. To determine whether dysmorphism of urinary red blood cells was a significant predictor of glomerular disease we compared the number of dysmorphic cells in the urine of patients with biopsy proven glomerulonephritis (GN), before and immediately after renal biopsy. We also compared the number of dysmorphic cells in patients with glomerulonephritis to those with lower urinary tract bleeding. Renal biopsy caused significant dysmorphic hematuria, indicating that dysmorphism suggests renal rather than glomerular bleeding. Although patients with GN had significantly more dysmorphic urinary RBC when compared to those with lower tract urinary bleeding, the overlap was such that one could only be confident of renal hematuria if they accounted for greater than 75% of the total number of RBC. Non renal hematuria is present if number of dysmorphic cells is less than 17% of total RBC. Thus dysmorphism of urinary RBC is a useful diagnostic tool, but only if strict criteria established for each laboratory are adhered to.