The aim of this review was to investigate the association between total and domain-specific physical activity (PA) and non-specific low back pain (LBP) in adults. Seven databases were searched for cohort and cross-sectional studies. Pooled estimates of the association of medium and high levels PA and LBP, using the generic inverse-variance method with fixed- and random-effects models were calculated. Twenty-four studies (15 cohort and nine cross-sectional; 95,796 participants) were included. The pooled fully adjusted risk ratios (RR) from cohort studies comparing medium with lowest activity levels were 0.90 (95%CI 0.85 to 0.96) for total PA, and 0.90 (95%CI 0.85 to 0.96) for leisure-time PA (LTPA). The pooled RR comparing highest with lowest activity levels were 1.00 (95%CI 0.92 to 1.08) for total PA, and 1.01 (95%CI 0.93 to 1.10) for LTPA. The pooled fully adjusted odds ratios (OR) from cross-sectional studies comparing medium with lowest activity levels were 0.93 (95%CI 0.65 to 1.32) for total PA, and 0.77 (95%CI 0.62 to 0.96) for LTPA. The pooled OR comparing highest with lowest activity levels were 1.05 (95%CI 0.89 to 1.23) for total PA, and 0.85 (95%CI 0.79 to 0.93) for LTPA. PA seems to be associated with lower prevalence of LBP.