This research examines why positively framed messages work more effectively than negatively framed messages in product advertising by establishing an affect priming process model. Findings from Experiment 1 showed that positively framed ad messages evoked higher levels of positive affect and lower levels of negative affect than did negatively framed ad messages. Accordingly, positively framed ad messages generated more favorable ratings on ad believability, ad liking, and brand attitudes. Most importantly, this research demonstrated the process by which frame-evoked affect exerted influence on brand attitudes via its impacts on priming affect-congruent cognitive responses. Experiment 1 also found that positively framed ads encouraged participants to be attentive to and elaborate on messages more so than negatively framed ads. Findings from Experiment 2 further showed that ad framing effects were moderated by the type of product attributes being featured. © 2008 Wiley Periodicals, Inc.