To estimate the potential of small doses of sarin (types I and II) and soman to cause delayed neuropathic effects, 400, 200, 61, and 0 micrograms/kg of sarin-I, 280, 140, 70, and 0 micrograms/kg of sarin-II, and 14.2, 7.1, 3.5, and 0 micrograms/kg of soman by gavage were compared with 510 mg/kg tri-o-cresyl phosphate (TOCP) in 14- to 18-month-old SPF white leghorn hens (4/dose) protected with atropine (100 mg/kg). The neuropathy target esterase (NTE) activity 24 hr after dosing was determined in brain, spinal cord, and lymphocytes and in plasma and brain for cholinesterase and carboxylesterase. None of the compounds showed statistically significant NTE decreases. Sarin-II showed a dose-related trend in the lymphocyte NTE (to 33% of control at 280 micrograms/kg), suggesting that longer exposure to lower doses might cause a cumulative neurotoxic insult. All of the agents decreased the activity of plasma and brain cholinesterase and carboxylesterase. Using more than 70% inhibition of brain NTE as a biochemical predictor of delayed neuropathy, sarin and soman appear unable to cause delayed neuropathy at nonlethal doses within this protocol.