High comorbidity between schizophrenia and tobacco addiction has been well established. Explanatory theories include nicotine as a cognitive enhancer ameliorating symptoms of schizophrenia and underlying shared substrates increasing susceptibility to addiction in these individuals. To test these non-mutually exclusive theories, the maternal immune activation (MIA) model was utilized. To this end, pregnant Sprague Dawley rats were subcutaneously injected with a bacterial endotoxin, lipopolysaccharide (0.5 mg/kg), on gestation days 10 and 11. Selective attention and working memory in adult male offspring were subsequently assessed using the latent inhibition and delayed non-matching to sample paradigms both before and after nicotine or saline self-administration. MIA led to deficits in both latent inhibition and delayed non-matching to sample in male offspring. Further, these animals showed a small but significantly increased responding for nicotine during self-administration acquisition, although there was no difference in dose-response effect or in progressive ratio testing. However, nicotine, but not saline self-administration, significantly ameliorated the cognitive deficits induced by MIA. While the male offspring of mothers prenatally exposed to lipopolysaccharide was only slightly more sensitive to the reinforcing effects of nicotine, after self-administration, the MIA-induced cognitive deficits significantly improved. These data lend support for the self-medication hypothesis of schizophrenia.