The ability of five organophosphorus nerve agents (tabun, sarin, soman, GF, and VX) to produce brain seizures and the effectiveness of atropine as an anticonvulsant treatment against these nerve agents were studied in two different animal models--the rat and guinea pig. All animals were implanted with cortical electrodes for EEG recordings. Five minutes after the start of nerve agent-induced EEG seizures, animals were treated intramuscularly (IM) with different doses of atropine sulfate and observed for seizure termination. The anticonvulsant ED50 of atropine sulfate for termination of seizures induced by each nerve agent was calculated and compared. In the rat model, selected oximes were administered either before, concurrent with, or following challenge with a 1.6 x LD50 dose of a given nerve agent to maximize seizure development with certain agent/oxime combinations. The choice and the timing of oxime administration significantly effected the incidence of seizure development by different nerve agents. When oxime administration did not effect seizure development (tabun, soman) the anticonvulsant ED50 for atropine sulfate was the same, regardless of the nerve agent used to elicit the seizure. When oxime administration reduced the incidence of seizure occurrence (sarin, GF, VX), the anticonvulsant ED50 dose of atropine sulfate for a nerve agent was lower. In the guinea pig model, animals were pretreated with pyridostigmine prior to challenge with 2 x LD50 of a given agent, and treated 1 min later with atropine sulfate (2 mg/kg) and 2-PAM (25 mg/kg). Under these conditions, the incidence, latency of seizure development, and anticonvulsant ED50s of atropine for soman-, tabun-, and GF-elicited seizures were virtually identical. With sarin, although the latency of seizure development was the same as with soman, tabun, and GF, seizures occurred with a lower incidence, and the anticonvulsant ED50 of atropine was lower. With VX, the latency of seizure development was notably longer, while the incidence of seizure development and anticonvulsant ED50 of atropine were significantly lower than with soman, tabun, or GF. In both models, a lower incidence of seizure development predicted a lower anticonvulsant dose of atropine. In the rat, the incidence of seizure development and the anticonvulsant effectiveness of atropine was highly dependent on the oxime used. In the guinea pig, higher doses of atropine sulfate were required to control soman-, tabun-, or GF-induced seizures, perhaps reflecting the lower cholinesterase reactivating ability of 2-PAM against these agents.