Background and objective: The effect of exercise in prevention of low back and pelvic girdle pain during pregnancy is uncertain. This study aimed to assess the effect of exercise on low back pain, pelvic girdle pain and associated sick leave. Databases and data treatment: Literature searches were conducted in PubMed, EMBASE, Cochrane Library, Google Scholar, ResearchGate and ClinicalTrials.gov databases from their inception through May 2017. Randomized controlled trials (RCTs) were eligible for inclusion in the review if they compared an exercise intervention with usual daily activities and at least some of the participants were free from low back pain and/or pelvic girdle pain at baseline. Methodological quality of included studies was evaluated using the Cochrane Collaboration's tool. A random-effects meta-analysis was performed, and heterogeneity and publication bias were assessed. Results: Eleven randomized controlled trials (2347 pregnant women) qualified for meta-analyses. Exercise reduced the risk of low back pain in pregnancy by 9% (pooled risk ratio (RR) = 0.91, 95% CI 0.83-0.99, I 2 = 0%, seven trials, N = 1175), whereas it had no protective effect on pelvic girdle pain (RR = 0.99, CI 0.81-1.21, I 2 = 0%, four RCTs, N = 565) or lumbopelvic pain (RR = 0.96, CI 0.90-1.02, I 2 = 0%, eight RCTs, N = 1737). Furthermore, exercise prevented new episodes of sick leave due to lumbopelvic pain (RR = 0.79, CI 0.64-0.99, I 2 = 0%, three RCTs, N = 1168). There was no evidence of publication bias. Conclusion: Exercise appears to reduce the risk of low back pain in pregnant women, and sick leave because of lumbopelvic pain, but there is no clear evidence for an effect on pelvic girdle pain. Significance: Exercise has a small protective effect against low back pain during pregnancy.