AIMS: To develop and validate a rapid and economical semiautomated approach to the measurement of immunostainable tissue components which is applicable to routine diagnostic practice. To apply this approach to the measurement of macrophages in renal biopsy specimens in nephrotic states, as protein in the renal tubules may induce macrophage infiltration, and the morphology of macrophages in tissue sections does not lend itself to cell counting. METHODS: Macrophages were identified by immunostaining with a pan-macrophage marker, followed by digital image capture and analysis using a macro procedure written for the freeware image analysis program NIH-Image. RESULTS: The method was rapid, robust and accurate to within the limits imposed by sampling error inherent in the use of small needle biopsy specimens. Very few macrophages are found in normal kidney (mean volume fraction (+/- 95% confidence limits) 0.04% (0.02%)) but infiltration of macrophages was detected in minimal change nephropathy (0.29% (0.12%)) and in membranous glomerulonephritis (0.42% (0.11%)). A statistically significant correlation was found between macrophage volume fraction and weight of proteinuria in minimal change nephropathy but not in membranous glomerulonephritis. Correlations were found in both diseases between macrophage volume fraction and serum creatinine at time of biopsy. CONCLUSIONS: The equipment is inexpensive and measurement takes less than one minute per biopsy specimen. The results indicate that macrophage infiltration is part of the pathological process in minimal change nephropathy and membranous glomerulonephritis. The correlation with creatinine at time of biopsy suggests that renal impairment in minimal change nephropathy may result from infiltration by immunologically active cells and not merely from haemodynamic changes in nephrons. However, the correlation is not close, indicating that the relation between macrophage infiltration and disease severity is not a simple one.