Three-dimensional morphometric analysis of segmental glomerulosclerosis in the rat. In idiopathic nephrotic syndrome, and in experimental models of nephrosis, changes of visceral epithelial cells involve the entire glomerular population while segmental sclerotic changes are reported to affect only a certain number of glomeruli. Because conventional determination of the percentage of glomeruli affected by sclerotic lesions is usually based on the examination of randomly selected sections, we wondered whether glomeruli appearing normal in a given section could be affected by sclerosis in other regions of the capillary tuft (CT). To assess the real incidence and the spatial extension of sclerotic changes at the level of single glomerulus, we used serial-section morphological analysis to measure the volume of the glomerulus and that of sclerosis lesions. In glomeruli from control rats and in glomeruli from adriamycin (ADR) treated rats surface area of Bowman's capsule (BC), CT and sclerotic regions were measured using stereology techniques in all the consecutive sections containing each individual glomerulus, and corresponding volumes were then calculated. Mean volume of BC and CT were not significantly different between control and ADR rats (0.71 ± 0.03 and 0.53 ± 0.03 vs. 0.76 ± 0.04 and 0.53 ± 0.02 µm³ × 10^-6, respectively). The distribution of glomerular volume parameters in the ADR rats were more spread out than in control animals, indicating that some glomeruli became smaller while other became larger. No sclerotic changes were found in control rats, while in the three ADR rats 94, 90 and 92% of glomeruli, respectively, were affected by sclerotic lesions. Conventional counting of glomeruli affected by sclerosis in single sections (on average 39, 27 and 23%) highly underestimated the percentage of sclerotic glomeruli. Despite sclerotic lesions involved almost entirely in the glomerular population, they were limited to a small fraction of the CT volume (on average 3.5, 2.2 and 2.6% in each animal). Our results indicate that: 1) adequate quantification of the percentage of glomeruli affected by sclerosis can be made only with three-dimensional morphologic analysis of the entire glomerulus, as glomeruli appearing normal in a given section can show sclerosis changes in regions far from the section of examination; and 2) in the model of ADR nephrosis the initiation of sclerotic processes simultaneously affects almost the entire glomerular population.