Physical activity and exercise messages are created by everyone from un(or mis)informed 'influencers' to public health and not for profit organizations, which results in an increasingly complex messaging landscape. Case in point: social media did not exist when I started my research career. This talk will give an overview of a research program that uses dual-processing theories to study the effects of physical activity messages on people's automatic and reasoned cognitions, and behaviour. Some of the first questions asked were who might attend to messaging elements and what those elements might be. This progressed to examining which associations may be activated by physical activity-related images or words, and then to the application of that knowledge to try to create more effective messages. Most recently, for example, we have looked at physical activity messages targeted at women with heart disease, the effects of nonstigmatizing images of people living with obesity being active, and the believability and effects of fitspo. The take home message? If you hope to persuade your target audience, don't assume that people are thinking about what you're trying to say and be aware (and wary) of other messages that may compete for attention.