Moderate‐certainty evidence shows that cannabis and cannabinoids probably lead to a reported reduction of 30% in spasticity in people with refractory spasticity in varying stages of multiple sclerosis (454 vs 249 per 1000 people; all results on average). More people also experienced a pain reduction of 50% or greater with cannabis and cannabinoids when compared with placebo, but the evidence was of very low certainty (458 vs 167 per 1000 people). Moderate‐certainty evidence shows that the use of cannabis and cannabinoids probably leads to a reported improvement in the Patient Global Impression of Change scale (268 vs 169 per 1000 people). However, low‐certainty evidence suggests that cannabis and cannabinoids might increase the risk of specific adverse events related to the nervous system (cognitive dysfunction, dizziness, somnolence, and headache; 524 vs 297 per 1000 people) and psychiatric disorders (confusion, paranoia, and psychosis; 127 vs 70 per 1000 people). There was little to no difference between treatment groups in serious adverse events, reported drug tolerance, and health‐related quality of life (very low‐ to low‐certainty evidence).