Objective: The present study tested whether taking into account both the reflective and the impulsive processes of physical activity (PA) is helpful in understanding how, and for whom, PA-promoting messages will be (in)effective in changing behavior. Method: Participants (N ! 101) were presented with a persuasive message promoting either PA (experimental condition) or healthy eating (control condition). Reflective intentions to be physically active were assessed both at baseline and after exposure to the message. Impulsive approach tendencies toward PA (IAPA) and sedentary behaviors (IASB) were assessed using a manikin task. The main outcome variable was accelerometer-assessed free time spent in moderate to vigorous physical activity (MVPA) over 1 week after exposure to the message. Results: Although the PA-promoting message had no direct effect on MVPA, the results showed that (a) this message increased intentions to practice PA, notably among participants with low or moderate (but not high) baseline intentions; (b) objective MVPA was positively predicted by postmessage PA intentions and IAPA, and negatively predicted by IASB; and (c) postmessage PA intentions predicted MVPA for individuals with low or moderate (but not high) IASB. A follow-up moderated mediation analysis corroborated these earlier results, showing that PA-promoting messages positively predicted MVPA through postmessage intentions only among individuals with low or moderate baseline intentions and low or moderate IASB. Conclusions: By identifying 2 boundary conditions, this study revealed important insights in explaining when PA-promoting messages will be effective to predict objective MVPA and when they will not.