Aiming to evaluate the effects of community-based interventions on the physical activity (PA) levels in adults, in December, 2017 a systematic review was developed in five electronic databases (Pubmed, Physical Education Index, Scopus, Sportdiscus e Web of Science) and in reference lists, looking for community-based randomised controlled trials (RCT) with interventions aimed at PA increasing levels assessed by objective methods in samples of adults. Risk of bias was assessed using an adapted version of the EPHPP instrument. Of 3,223 articles initially identified, nine composed the descriptive synthesis. Most interventions focused on inactive people and women. In view of the heterogeneity between the PA indicators, a descriptive synthesis of the data was chosen, thus, considering number of steps per day, four of five interventions showed statistically positive results. In regard of moderate-to-vigorous PA levels, two of five interventions showed potential results. Most of interventions were based on educational activities as practices of counselling, home visits and telephone support. Articles with moderate risk of bias were the majority in four of the five domains assessed. In conclusion, most of the studies included showed the potential of community-based interventions to increase the number of steps per day. It is important that future studies discuss operational strategies of interventions, such as counselling and home visits