Abstract Objective Physical activity (PA) improves health outcomes in people with schizophrenia. It is unclear how much PA people with schizophrenia undertake and what influences PA participation. We conducted a meta-analysis to investigate PA levels and predictors in people with schizophrenia. Method Major databases were searched from inception till 02/2016 for articles measuring PA (self-report questionnaire (SRQ) or objective measure (e.g. accelerometer)) in people with schizophrenia, including first episode psychosis (FEP). A random effects meta-analysis and meta-regression analysis were conducted. Results 35 studies representing 3453 individuals with schizophrenia (40.0 years; 64.0% male) were included. Engagement in light PA was 80.44 min (95% CI 68.32–92.52, n = 2658), 47.1 min moderate-vigorous PA (95% CI 31.5–62.8, n = 559) and 1.05 min (95% CI 0.48–1.62, n = 2533) vigorous PA per day. People with schizophrenia engaged in significantly less moderate (hedges g = − 0.45, 95% CI − 0.79 to − 0.1, p = 0.01) and vigorous PA (g = − 0.4, 95% CI − 0.60 to − 0.18) versus controls. Higher light to moderate, but lower vigorous PA levels were observed in outpatients and in studies utilizing objective measures versus SRQ. 56.6% (95% CI 45.8–66.8, studies = 12) met the recommended 150 min of moderate physical activity per week. Depressive symptoms and older age were associated with less vigorous PA in meta-regression analyses. Conclusions Our data confirm that people with schizophrenia engage in significantly less moderate and vigorous PA versus controls. Interventions aiming to increase PA, regardless of intensity are indicated for people with schizophrenia, while specifically increasing moderate-vigorous PA should be a priority given the established health benefits.