Ifosfamide (IF) is an alkylating cytostatic derived from nitrogen mustard. In addition to its well-known urotoxic effects (hemorrhagic cystitis), several cases of Fanconi syndrome following IF therapy have been reported. No information is available concerning the pathomechanisms of this tubulotoxicity. We used the permanent renal epithelial cell line LLC-PK1 in order to investigate whether major metabolites of IF (i.e. 4-OH-IF, acrolein and chloracetyldehyde) induce the transport defects most frequently detected after IF therapy in vivo. LLC-PK1 cells of passages 162-177, grown in plastic culture dishes, were used in a confluent state. Sodium-dependent and independent fluxes of l-[3H]alanine and of D-[3H] glucose were determined by standard techniques. Activities of marker enzymes of apical and basolateral membranes, of mitochondria and of endoplasmic reticulum were determined in cell homogenates. IF itself has no detectable effect on fluxes of /-alanine and D-glucose in LLC-PK, cells. The IF metabolite 4-OOH-IF induces a clear inhibition of sodium-dependent fluxes of both substrates after a 24-hour exposure of cells to 100 µmol/l of 4-OOH-IF. Chloracetaldehyde induces a biphasic response of sodium-dependent fluxes of l-alanine with increased uptake rates at low concentrations ( < 200 µmol/l) and with a short incubation time, while higher concentrations and long exposure of the cells leads to a reduction in sodium coupled transport. Glucose transport is affected in a comparable way, however, in contrast to alanine transport, chloracetaldehyde also stimulates sodium-independent fluxes of glucose. Acrolein is the most toxic substance tested. It severely damages cell monolayers at concentrations beyond 75 µmol/l. Sodium-coupled glucose and alanine transport is inhibited by acrolein at concentrations higher than 50 µmol/l. Sodium-coupled glucose transport is more sensitive to all metabolites tested than alanine transport. While acrolein strongly affects both transport systems, marker enzymes of the apical plasma membrane, i.e. alkaline phosphatase and leucine aminopeptidase, are not significantly inhibited, suggesting a specificity of the toxic effect for the transport proteins. We conclude that LLC-PK1 cells represent a good model for further investigation of the pathogenesis of Fanconi syndrome after IF therapy. Sodium-dependent transport systems are more sensitive to acrolein than other cell surface proteins.