The predominant hypothesis for cyclosporine-induced acute renal failure is postulated to be prerenal vasoconstriction with concomitant hemodynamic changes; an alternate hypothesis, however, may be that cyclosporine (CsA) affects intrarenal processes, i.e., direct renal parenchymal cell injury. However, reports on this direct effect of CsA on renal parenchymal cells are contradictory. Therefore, the purpose of this study was to address whether CsA is directly toxic to renal parenchymal cells in a primary culture system of rat renal cortical epithelial cells. The cytotoxicity of Sandimmune, the commercial form of CsA in a polyoxyethylated castor oil vehicle (Cremophor), CsA without vehicle, and the Cremophor vehicle was assessed by plasma membrane integrity (lactate dehydrogenase leakage), mitochondrial metabolic activity [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction], and gross morphology (phase-contrast microscopy). The cytotoxicity of Sandimmune was also assessed by lysosomal activity (neutral red uptake), by proximal tubular enzyme activity (alkaline phosphatase), and by three fluorescent probes using a multiwell scanner. The three fluorescent probes were propidium iodide which stains nuclei of nonviable cells; bis-carboxyethyl-carboxyfluorescein which is retained by viable cells; and rhodamine 123, which assesses mitochondrial membrane potential. The results of this study demonstrated that Sandimmune caused dose- (10, 25, and 50 microM) and time- (12, 24, and 48 hr) dependent cytotoxicity, while Cremophor caused cytotoxicity only at high concentrations and long incubations. We conclude that (1) CsA is directly toxic to renal parenchymal cells in vitro and this system potentially represents a sensitive model for further mechanistic studies; (2) CsA plus vehicle (Sandimmune) was more cytotoxic to renal cells than CsA alone (without the polyoxyethylated castor oil vehicle).