Two studies investigated the impact of affective and cognitive messages compared to a no-message control on self-reported exercise. Students (Study 1, N = 383 and Study 2, N = 197) were randomly allocated to one of the three conditions (control - no message, affective message or cognitive message). Participants completed questionnaire measures tapping components of the theory of planned behaviour in relation to exercise and reported their level of exercise (3 weeks later). In Study 2, measures of need for affect (NFA) and need for cognition (NFC) were also completed. Results showed that affective messages consistently produced greater increases in self-reported level of exercise than the other conditions. In both studies, this effect was partly mediated by affective attitude change. Study 2 indicated these effects to be significantly stronger among those high in NFA or low in NFC. These findings indicate the value of affective messages that target affective attitudes in changing exercise behaviour.