The aim of this study is to examine the effects of treatment with varenicline, a partial agonist at the a4b2 and full agonist at the a7 nicotine acetylcholine receptor, on cognitive impairments in people with schizophrenia. In all, 120 clinically stable people with schizophrenia participated in randomized, double-blind, placebo-controlled 8-week trial. Antipsychotic and concomitant medication doses remained fixed throughout the study. Varenicline was titrated up to 1 mg twice daily for weeks 2-8. Neuropsychological, clinical, and safety assessments were administered at baseline and weeks 1, 2, 4, and 8. In the primary analyses of neurocognitive differences at week 8, no varenicline-placebo differences were significant. In secondary longitudinal analyses, varenicline improved compared with placebo on the Digital Symbol Substitution Test (p ¼ 0.013) and the Wisconsin Card Sorting Test non-perseverative errors (p ¼ 0.043). Some treatment effects were different between smokers and non-smokers. In smokers, Continuous Performance Test hit reaction time (p ¼ 0.008) and Stroop Interference (p ¼ 0.004) were reduced for varenicline compared with placebo, while there were no treatment differences in non-smokers. No significant treatment main effects or interactions were noted for total scores on the Positive and Negative Syndrome Scale or the Scale for the Assessment for Negative Symptoms. Our findings suggest beneficial effects of adjunctive varenicline treatment with antipsychotics for some cognitive impairments in people with schizophrenia. In some cases, effects of treatment varied between smokers and non-smokers. Further study is required to assess the functional significance of these changes.