The usefulness of the morphologic examination of urinary red cells was first described by Birch and Fairley, who reported that the dysmorphic red cells(morphologically variable) were the marker for glomerular bleeding and the isomorphic red cells(morphologically uniform) for non-glomerular bleeding. They also noted that healthy individuals had dysmorphic cells, indicating a glomerular source. The relation between red blood cell morphology and the origin of hematuria has been confirmed subsequently by numerous clinical studies. Although the reports varied on the ratio(10-100%) of dysmorphic cells, their findings have been supported by many investigators. In 1991 Köhler noted that acanthocytes(AC) were the most characteristic red cell type for glomerular bleeding and that acanthocyturia > or = 5% was a good predictive marker for glomerular bleeding. Moreover the red cell was easily recognized. On the other hand, in 1993 Fairley and Birch reported that in glomerulonephritis, erythrocytes varied markedly in size, shape, and hemoglobin content, and that the pattern of morphology indicates the source of bleeding, not the morphology of individual cells. Both observations were confirmed in our in vitro experiments. To clarify the causative mechanism of acanthocyturia, the normal washed erythrocytes were successively exposed to two kinds of solutions simulating conditions in the distal tubules(105 mOsm, pH 5.5, NaCl 37 mmol/l, KH2PO4 1.0 mmol/l, Urea 0.9 g/l) and the collecting tubules(a mixture of NaCl and KH2PO4 solutions with the osmolality between 299 and 1192 mOsm). Various degrees of hemolytic process appeared in the first solution. However, AC appeared only in the second solutions with 9 mmol/l of KH2PO4 and 390 mOsm or more osmolality or with 725 mOsm and 4.1 mmol/l or more of KH2PO4. This result supported the nephron passage theory in the formation of AC in glomerular diseases.