Graphic warnings (GWs) on cigarette packs are widely used internationally with the aim of reducing smoking behavior. In the current study, we investigated whether GWs influence implicit evaluations of smoking, a potential moderator of smoking behavior, as measured with an Implicit Association Test (IAT). Results showed that viewing a GW did not produce more negative implicit evaluations of smoking for daily smokers, occasional smokers, or non-smokers, compared to viewing a text-only warning. If anything, effects were in the direction of evaluations of smoking being more positive after certain participants (i.e., daily and occasional smokers) viewed a GW. We also did not find any beneficial effects of GWs on explicit evaluations of smoking. These results contrast with the observation that non-smokers and occasional smokers (but not daily smokers) believed that GWs would be more effective than the text-only warnings. We discuss implications and limitations of these findings and provide recommendations for improving the effectiveness of cigarette pack warnings on implicit evaluations.