The long-term treatment of Parkinson disease (PD) may be complicated by the development of levodopa-induced dyskinesia. Clinical and animal model data support the view that modulation of cannabinoid function may exert an antidyskinetic effect. The authors conducted a randomized, double-blind, placebo-controlled crossover trial to examine the hypothesis that cannabis may have a beneficial effect on dyskinesia in PD. A 4-week dose escalation study was performed to assess the safety and tolerability of cannabis in six PD patients with levodopa-induced dyskinesia. Then a randomized placebo-controlled crossover study (RCT) was performed, in which 19 PD patients were randomized to receive oral cannabis extract followed by placebo or vice versa. Each treatment phase lasted for 4 weeks with an intervening 2-week washout phase. The primary outcome measure was a change in Unified Parkinson's Disease Rating Scale (UPDRS) (items 32 to 34) dyskinesia score. Secondary outcome measures included the Rush scale, Bain scale, tablet arm drawing task, and total UPDRS score following a levodopa challenge, as well as patient-completed measures of a dyskinesia activities of daily living (ADL) scale, the PDQ-39, on-off diaries, and a range of category rating scales. Seventeen patients completed the RCT. Cannabis was well tolerated, and had no pro- or antiparkinsonian action. There was no evidence for a treatment effect on levodopa-induced dyskinesia as assessed by the UPDRS, or any of the secondary outcome measures. Orally administered cannabis extract resulted in no objective or subjective improvement in dyskinesias or parkinsonism.