The current studies were designed to test two hypotheses: (1) differences in steady-state reduced glutathione levels are responsible for subcompartment differences in susceptibility to acute ozone injury, and (2) elevation of reduced glutathione concentrations accounts for the tolerance to further injury produced by repeated ozone exposure. Glutathione was measured in well-defined subcompartments of the lung of both rats and monkeys to compare alterations occurring in both target (distal trachea and terminal bronchiole) and nontarget areas (lobar bronchus, major daughter, minor daughter bronchus, and parenchyma) of the lung in species that differ in sensitivity to ozone exposure (rat is less susceptible than monkey). Glutathione concentrations were decreased in trachea of rats exposed to 0.4 ppm ozone for 2 h and increased in lobar bronchus and distal bronchiole after 2 h exposure at 1 ppm. In monkey, glutathione levels in most subcompartments were not altered by either 0.4 or 1.0 ppm ozone exposure for 2 h. The exceptions were the major daughter subcompartment (200% of control at 0.4 ppm exposure) and the distal bronchiole (55% of control at 1 ppm exposure). Ninety day ozone exposures (6 h/day x 5 days/week) in rats produced an elevation in glutathione (164% of control value) only in distal bronchiole at the 1 ppm exposure level. In a similar manner, glutathione levels in the distal bronchiole of monkeys exposed for 90 days to 1 ppm O3 were 165% of the corresponding control values. These results suggest the following: glutathione levels in target and nontarget areas of the lung and in susceptible versus less susceptible species are not the primary determinant in the differences observed in ozone toxicity; the response of lung subcompartments to short-term ozone exposure varied depending on airway subcompartment and species; increased glutathione levels may be one reason for adaptation of some airway epithelial cells from rats and monkeys exposed to O3 for long periods; and use of well-defined segments of the lung provides a means of assessing changes in target areas of the lung without dilution from nontarget areas.