Objectives: To examine the impact of an exercise imagery intervention (vs. an attention control) on the primary outcome of exercise identity, and secondary outcomes of behaviour and fitness, of female initiates during their involvement in a structured 8-week exercise programme (with two follow-up assessments). Design: Randomized experimental design with pre and post measures. Method: Female exercise initiates (N = 145) were randomly assigned into an imagery or control condition. Each participant was given an individualised, cardiovascular programme to follow for eight weeks. Participants listened to tailored imagery scripts lasting 3–5 min each week during the 8-week programme and at week 0, 5, 9, 18, and 36 participants competed assessments. Results: Findings revealed significant improvements over time for exercise identity, exercise behaviour, and fitness. In addition, a significant group x time interaction effect was observed at week 9 for identity in favour of the exercise imagery condition. Conclusion: Given exercise identity appears to serve as an important construct in the promotion of exercise adherence (Strachan et al., 2013. Understanding variations in exercise-identity strength through identity theory and self-determination theory. International Journal of Sport and Exercise Psychology, 11(3), 273–285. doi:10.1080/1612197X.2013.749005), results of the imagery intervention in the current study are very promising; exercise imagery can be used to improve exercise role identity.