OBJECTIVE: To determine if habitual physical activity could be increased in people with cerebral palsy. DATA SOURCES: We searched electronic databases until February 2010 using key words related to concepts of cerebral palsy and physical activity. This search was supplemented with citation tracking. METHODS: Studies had to include participants with cerebral palsy who have habitual physical activity measured over at least one day after a therapy intervention. Two reviewers independently assessed study quality with the PEDro scale (quantitative studies) and Critical Appraisal Checklist for Qualitative Research (qualitative studies). For quantitative studies standardized mean differences were calculated and meta-analysis conducted. Qualitative data were synthesized thematically. RESULTS: Three randomized controlled trials (96 participants) and two qualitative studies (21 participants) were reviewed. Four studies evaluated exercise programmes, and one study an online educational and support programme. Meta-analysis showed that exercise programmes could increase habitual physical activity (δ = 1.0; 95% confidence interval (CI) 0.28 to 1.72). This result was reinforced by reports of increased daily activity in two qualitative studies. The online programme increased weekly minutes of moderate to vigorous physical activity (d = 0.81; 95% CI 0.17 to 1.45), and weekly step counts (d = 0.62; 95% CI 0.0 to 1.25). Positive effects were not maintained after programmes stopped. There was insufficient evidence to determine if demographic factors or programme characteristics, such as intensity and setting, were associated with outcomes. CONCLUSION: Preliminary evidence suggests that exercise programmes and online support programmes can increase habitual physical activity in people with cerebral palsy, but effects are not maintained when programmes stop.