OBJECTIVES: Cannabis has been proposed as a potential treatment for Parkinson's disease (PD) due to its neuroprotective benefits. However, there has been no rigorous review of preclinical studies to evaluate any potential treatment effect. This systematic review was undertaken to provide evidence in support or against a treatment effect of cannabinoids in animal models of PD. METHODS: Databases were searched for any controlled comparative studies that assessed the effects of any cannabinoid, cannabinoid-based treatment or endocannabinoid transport blocker on behavioural symptoms in PD animal models. RESULTS: A total of 41 studies were identified to have met the criteria for this review. 14 of these studies were included in meta-analyses of rotarod, pole and open field tests. Meta-analysis of rotarod tests showed a weighted mean difference of 31.63 s for cannabinoid-treated group compared with control. Meta-analysis of pole tests also showed a positive treatment effect, evidenced by a weighted mean difference of -1.51 s for cannabinoid treat group compared with control. However, meta-analysis of open field test demonstrated a standardised mean difference of only 0.36 indicating no benefit. CONCLUSION: This review demonstrates cannabinoid treatment effects in alleviating motor symptoms of PD animal models and supports the conduct of clinical trials of cannabis in PD population. However, there is no guarantee of successful clinical translation of this outcome because of the many variables that might have affected the results, such as the prevalent unclear and high risk of bias, the different study methods, PD animal models and cannabinoids used.