AbstractThe purpose of this pilot study was to indicate, for low‐level exposure of conscious guinea pigs and marmoset monkeys to sarin vapour in air, the lowest‐observable‐adverse‐effect level (LOAEL) of sarin for miosis. This is the concentration × time (C·t) value (t = 5 h) of exposure at which miosis becomes significant.The ratio of pupil and iris diameters, measured on digital photographs taken on‐line during exposure, was calculated as a measure for miosis. The exposure concentrations were in the range 7–150 µg m−3 and the exposure times needed to achieve significant miosis were in the range 10–300 min. Both vehicle‐ and pyridostigmine‐pretreated animals were used in the experiments. The latter pretreatment resulted in ca. 30% inhibition of erythrocyte acetylcholinesterase in both species.In vehicle‐pretreated guinea pigs and marmosets the pupil size was decreased significantly (P < 0.05) at sarin doses of 1.8 ± 0.3 and 2.5 ± 0.8 mg min m−3, respectively. In pyridostigmine‐pretreated guinea pigs and marmosets the pupil size was affected significantly (P < 0.05) at 1.8 ± 0.5 and 3.0 ± 0.8 mg min m−3, respectively. Evidently there is no significant influence of pyridostigmine pretreatment on the LOAEL. These data were addressed in light of the recommended occupational and detection limits for sarin vapour in air. It was concluded that miosis will occur during low‐level sarin exposure at levels that are not detectable by the currently fielded alarm systems, assuming that humans are as sensitive for sarin vapour in air as guinea pigs and marmosets. Copyright © 2004 John Wiley & Sons, Ltd.