We evaluated the efficacy of pretreatment with phenytoin and phenobarbital to prevent seizures in mice given convulsive doses of theophylline. The control LD50 for theophylline was determined in 48 mice by intraperitoneal injections of increasing doses without anticonvulsant treatment. Anticonvulsant effects were determined in 105 additional mice pretreated with either phenytoin 30 mg/kg (n = 35), phenobarbital 35 mg/kg (n = 30), or phenobarbital 60 mg/kg (n = 40) one hour before theophylline administration. The theophylline LD50 (95% confidence interval) was 239 mg/kg (range, 229 to 248 mg/kg) for controls, 204 mg/kg (range, 194 to 214 mg/kg) for phenytoin, 305 mg/kg (range, 288 to 323 mg/kg) for low-dose phenobarbital, and 319 mg/kg (range, 307 to 331 mg/kg) for high-dose phenobarbital. Each LD50 differed significantly from control (P less than .05). The phenobarbital groups were significantly different from phenytoin (P less than .05) but not from each other. Theophylline serum concentrations were not significantly different among groups after adjustment for different doses. The mean +/- SEM time to seizure in minutes after adjustment for theophylline dose was 23.5 +/- 4.0 minutes for controls, 5.7 +/- 7.5 minutes for phenytoin, 44.1 +/- 7.1 minutes for low-dose phenobarbital, and 63.7 +/- 6.5 minutes for high-dose phenobarbital.(ABSTRACT TRUNCATED AT 250 WORDS)