Background: Venom immunotherapy can be initiated by different schedules, but randomized comparisons have not been performed. Objective: We aimed to compare the safety of 2 initiation schedules. Methods: Patients of any age with prior immediate generalized reactions to jack jumper ant (Myrmecia pilosula) stings were randomized to venom immunotherapy initiation by a semirush schedule over 10 visits (9 weeks) or an ultrarush schedule over 3 visits (2 weeks). In a concurrent treatment efficacy study, the target maintenance dose was randomized to either 50 μg or 100 μg. The primary outcome was the occurrence of 1 or more objective systemic reactions during venom immunotherapy initiation. Analyses were by intention to treat. We also assessed outcomes in patients who declined randomization. Results: Of 213 eligible patients, 93 were randomized to semirush (44 patients) or ultrarush (49 patients) initiation. Objective systemic reactions were more likely during ultrarush initiation (65% vs 29%; P <.001), as were severe reactions (12% vs 0%; P =.029). Times to maximal increases in venom-specific IgG 4 were no different between treatments, whereas the maximal increase in venom-specific IgE occurred earlier with ultrarush treatment. Similar differences between methods were observed in patients who declined randomization. One hundred seventy-eight patients were randomized to maintenance doses of either 50 μg (90 patients) or 100 μg (88 patients). The target maintenance dose had no effect on the primary outcome, but multiple-failure-per-subject analysis found that the 50 μg dose reduced the likelihood of reactions. Conclusion: Ultrarush initiation increases the risk of systemic reactions. A lower maintenance dose reduces the risk of repeated reactions, but the effect on treatment efficacy is unknown.