Acute exposure to phosgene, a toxic gas widely used in industrial processes, decreases resistance to bacteria in mice and rats and enhances susceptibility to B16 tumor cell challenge in mice. These effects appear to De due to impaired alveolar macrophage and natural killer (NK) cell activity, respectively. In this study effects of repeated phosgene exposures on bacterial infection and NK activity were determined. Rats were exposed for 4 or 12 wk, 6 h/day, 5 days/wk, to 0.1 or 0.2 ppm phosgene or 2 days/wk to 0.5 ppm and infected by aerosol with Streptococcus zooepidemicus immediately after the last exposure. An additional group was also infected after 4 wk of recovery following the 12-wk exposure regimens. Bronchoalveolar lavage (BAL) fluid was assessed 0, 6, and 24 h postinfection for bacteria and inflammatory cells. Differential cell counts in BAL and pulmonary NK activity were also determined in uninfected rats 18 h after the last exposure. All phosgene exposures impaired clearance of bacteria from the lungs and caused an increase in polymorphonuclear leukocytes (PMNs) in BAL of infected rats. Effects in the 0.5 ppm exposure group were greatest, and were significantly different from those in the 0.2 ppm exposure group, although the product of concentration x time was the same. BAL cell counts and bacterial clearance were normal in rats assessed 4 wk after the 12-wk phosgene exposures. Bacterial clearance and the PMN response to infection following repeated exposure were similar to those observed after a single exposure; that is, for these endpoints, effects due to repetitive exposure were neither additive nor attenuated. In contrast, NK activity was suppressed only at the 0.5 ppm level, and the magnitude of suppression was much less than that following acute exposure, suggesting that attenuation of this effect did occur with repeated exposure. The data indicate that susceptibility to streptococcal infection is a sensitive endpoint for phosgene toxicity following subchronic exposure.