STUDY DESIGN: Systematic review. OBJECTIVES: To evaluate the impact of cannabinoids on neurobehavioral outcomes in preclinical models of nontraumatic and traumatic spinal cord injury (SCI), with the aim of determining suitability for clinical trials involving SCI patients. METHODS: A systematic search was performed in MEDLINE and Embase databases, following registration with PROPSERO (CRD42019149671). Studies evaluating the impact of cannabinoids (agonists or antagonists) on neurobehavioral outcomes in preclinical models of nontraumatic and traumatic SCI were included. Data extracted from relevant studies, included sample characteristics, injury model, neurobehavioural outcomes assessed and study results. PRISMA guidelines were followed and the SYRCLE checklist was used to assess risk of bias. RESULTS: The search returned 8714 studies, 19 of which met our inclusion criteria. Sample sizes ranged from 23 to 390 animals. WIN 55,212-2 (n = 6) and AM 630 (n = 8) were the most used cannabinoid receptor agonist and antagonist respectively. Acute SCI models included traumatic injury (n = 16), ischaemia/reperfusion injury (n = 2), spinal cord cryoinjury (n = 1) and spinal cord ischaemia (n = 1). Assessment tools used assessed locomotor function, pain and anxiety. Cannabinoid receptor agonists resulted in statistically significant improvement in locomotor function in 9 out of 10 studies and pain outcomes in 6 out of 6 studies. CONCLUSION: Modulation of the endo-cannabinoid system has demonstrated significant improvement in both pain and locomotor function in pre-clinical SCI models; however, the risk of bias is unclear in all studies. These results may help to contextualise future translational clinical trials investigating whether cannabinoids can improve pain and locomotor function in SCI patients.