Introduction: One of the most common adverse effects of habitual cannabis use is hyperemesis—recurrent bouts of protracted vomiting, retching and abdominal pain superimposed on a baseline of daily nausea and anorexia. Largely anecdotal evidence supports the use of haloperidol, benzodiazepines or topical capsaicin over traditional antiemetics, yet little is known about the cause or optimal treatment of this newly recognized disorder. We report the results of one of the first clinical trials on so-called cannabis hyperemesis syndrome (NCT03056482). Methods: We approached adults with a working diagnosis of hyperemesis due to cannabis, provided they had ongoing emesis for >2 hours, a cyclic pattern of 3+ episodes in the last 2 years, and near daily use of cannabis by inhalation. We excluded those who were pregnant, deemed unreliable, or using opioids. Subjects provided written consent to be randomized during the index or any subsequent visit to either haloperidol (with a nested randomization to either 0.05 mg/kg or 0.1 mg/kg) or ondansetron 8 mg intravenously in a quadruple-blind fashion, and to be followed for 7 days. The primary outcome was the average reduction from baseline in abdominal pain and nausea (each measured on a 10-cm VAS) at 2 hours. While the original trial design allowed for crossover, the primary analysis used only the first treatment period since fewer than the prespecified threshold of 20% of subjects crossed over. Results: We enrolled 33 subjects, of whom 30 (16 men, 29+/-11 years old, using 1.5+/-0.9 g/day since age 19+/-2 years) were treated at least once (haloperidol 13, ondansetron 17). Haloperidol at either dose was superior to ondansetron (difference 2.3 cm [95%CI 0.6, 4.0]; p = 0.01), with similar improvements in both pain and nausea, as well as less rescue antiemetics (27% vs 61%; p = 0.04), and shorter time to ED departure (3.1+/-1.7 vs 5.6+/-4.5 hours; p = 0.03 Wilcoxon rank sum). There were two (haloperidol) vs six (ondansetron) return visits for ongoing nausea/vomiting, as well as two return visits for acute dystonia, both in the higher dose haloperidol group. Conclusion: Haloperidol is superior to ondansetron for the acute symptomatic treatment of patients with ongoing hyperemesis attributed to habitual cannabis use. The efficacy of this agent over ondansetron provides insight into the mechanism of this new disorder, now almost a daily diagnosis in many Canadian emergency departments.