Background: Physical activity (PA) levels of older adults living in a care setting are known to be very low. This is a significant health(care) problem, as regular PA has many health benefits also at advanced age. Research on automatic processes underlying PA behaviour in physically inactive older adults is yet non-existing. Since people are unconsciously influenced by people around them (i.e. by 'social norms') automatic processes could be used to promote PA. We developed an explorative intervention method to assess the effects of automatically processed (implicit) descriptive social norms ('What most people do') on behavioral intention and participation in PA offered in a local residential care setting. Methods: Forty-seven care clients met the inclusion criteria. Participants (response 45%; unaware of the intention of the research) were randomly assigned to an experimental (N = 10) or a control group (N = 11). The experimental group was exposed to photos and text heading on active peers (physically active implicit descriptive norm) using a draft newsletter article they were asked to comment on, whereas the control group was exposed to a newsletter with photos and text heading of inactive peers (physically inactive implicit descriptive norm). Subsequently, we tested (Fishers exact p < 0.10) whether this unaware exposure predicted intention (implicit and explicit) to participate in PA offered and organized by the care center (e.g. walking, gymnastics) and self-reported participation in organised PA at three months follow-up. Participants were debriefed later. Results: Mean age was 87 years (SD = 3.6; range 80-95) and 53% of the participants were male. At baseline, there were no significant differences in self-rated health and PA between the experimental and control group. Results indicated that implicit descriptive norm information was associated with implicit PA intention (p =.056, Fisher's exact test). No significant effects were found on explicit intention. At 3 months follow-up the experimental group self-reported 80% participation in PA versus 22% in the control group (Fisher's exact test p = 0.027). Conclusion: Implicit descriptive social norm information could indeed be a potentially effective way to encourage inactive older adults in residential care to engage in organized PA.