The central nervous system toxicity of cyclosporine, which is known to be neurotoxic clinically, was investigated in a rat model. Munich-Wistar rats were divided into 3 groups for a 2-week protocol. After baseline EEG and behavioral testing, group 1 (control) received a weight-adjusted volume of parenteral cyclosporine vehicle i.p., group 2 (low-dose) received 5 or 10 mg/kg/day i.p., and group 3 (high-dose) received 20 mg/kg/day i.p., Spontaneous behavior was observed, simple sensorimotor testing performed daily, and awake EEG's recorded 3 times per week. Four of 12 high-dose animals died during study, one after a witnessed tonic-clonic seizure, and two after recording of frankly epileptiform EEG's; there were no deaths in control or low-dose animals. Significant EEG abnormalities developed only at high-dose, with frankly epileptiform EEG's and/or seizures seen in 58±15% of these rats (P=0.005, different from controls by life-table analysis). Although some high-dose animals demonstrated hyperirritability and dystonic posturing, behavioral changes were subtle, and animals were often still or rocking slightly during recording of frankly epileptiform EEG's. Walking latency and alley escape behaviors were delayed in high-dose rats, the latter correlating with abnormal EEG's. Serum urea nitrogens were mildly elevated in high-dose animals, but serum creatinine, electrolytes, bilirubin, body magnesium stores, and blood pressure remained normal in all groups. Kidneys showed only mild vacuolation histologically. The brain showed only very focal cortical injury sites related to electrode placement, which did not correlate with EEG changes or mortality. These results suggest that there may be a direct effect of cyclosporine on the central nervous system. This model system should prove useful in defining mechanisms of cyclosporine-related neurotoxicity.