Purpose: To identify interventions employed to increase post-stroke physical activity, evaluate their efficacy, and identify the gaps in literature. Materials and methods: Randomized controlled trials published until March 2018 were searched in MEDLINE, PEDro, EMBASE, LILACS, and SCIELO databases. The quality of each study and overall quality of evidence were assessed using the PEDro and the GRADE scales. Results: Eighteen studies were included (good PEDro and very low GRADE-scores). In seven, the experimental groups showed significant increases in physical activity (aerobics, resistance, and home-based training; counseling, aerobics, resistance, and home-based training; electrical stimulation during walking; functional-task training; robot-assisted arm therapy; accelerometer-based feedback, and physical activity encouragement). In seven, there were no significant between-group differences (physical activity plan; stretching, use of toe-spreaders, standard treatment; counseling; circuit video-game; functional-task; counseling and cognitive training). The combined experimental and control groups showed significant declines in physical activity in one study (aerobic training or stretching) and increases in three others (aerobic, resistance or sham resistance training; stroke-with advice or only stroke-counseling; aerobic training, educational sessions, standard treatment, and coaching, or mobilization and standard treatment). A meta-analysis could not be performed, due to heterogeneity. Conclusions: Some interventions improved physical activity after stroke. However, the interpretability is limited.