Numerous preclinical studies show that acute cannabinoid administration impairs cognitive performance. Almost all of this research has employed cannabinoid injections, however, whereas smoking is the preferred route of cannabis administration in humans. The goal of these experiments was to systematically determine how acute exposure to cannabis smoke affects working memory performance in a rat model. Adult male (n = 15) and female (n = 16) Long-Evans rats were trained in a food-motivated delayed response working memory task. Prior to test sessions, rats were exposed to smoke generated by burning different numbers of cannabis or placebo cigarettes, using a within-subjects design. Exposure to cannabis smoke had no effect on male rats’ performance, but surprisingly, enhanced working memory accuracy in females, which tended to perform less accurately than males under baseline conditions. In addition, cannabis smoke enhanced working memory accuracy in a subgroup of male rats that performed comparably to the worst-performing females. Exposure to placebo smoke had no effect on performance, suggesting that the cannabinoid content of cannabis smoke was critical for its effects on working memory. Follow-up experiments showed that acute administration of either Δ9-tetrahydrocannabinol (0.0, 0.3, 1.0, 3.0 mg/kg) or the cannabinoid receptor type 1 antagonist rimonabant (0.0, 0.2, 0.6, 2.0 mg/kg) impaired working memory performance. These results indicate that differences in the route, timing, or dose of cannabinoid administration can yield distinct cognitive outcomes, and highlight the need for further investigation of this topic.