Background Onabotulinum toxin A (OnA) is a well-tolerated and effective treatment for chronic migraine (CM). However, based on research indications that incobotulinum toxin A (InA) would be equally effective, a Veterans Health Administration medical center mandated a two-year trial of InA as a more cost-effective alternative to OnA. Although InA is used for many similar indications as OnA, it is not Food and Drug Administration-approved for treating CM, and complications occurred in several patients with CM following this treatment change. We conducted this retrospective analysis to evaluate differences in the efficacy of OnA and InA and identify the reasons for the adverse effects of InA in some of these patients. Methods We performed a retrospective review of 42 patients who had been effectively treated with OnA and were then switched to InA. The differences between treatment responses to OnA and InA were assessed through the evaluation of pain on injection, number of headache days, and duration of action. Patients received injections at 10- to 13-week intervals. Those who reported severe pain on injection of InA were switched back to OnA. Results Severe burning pain on InA injection was reported by 38% of patients (nine males and seven females, i.e., a total of 16 patients out of 42 patients). One male patient reported the same degree of pain from both InA and OnA injections. A total of 66.7% of women with obesity and 83.3% of men with obesity or diabetes experienced severe pain on injection. Neither migraine suppression nor the duration of effect was significantly different between OnA and InA. Conclusions OnA is better tolerated than InA in the treatment of CM. InA appears to effectively suppress migraines, but some patients complain of a severe localized burning sensation during the injections. Some of these patients, all of whom were previously treated with OnA, requested to switch back to OnA. This suggested that InA is not equivalent to OnA in terms of tolerability and effectiveness. The present study found 2.38% of patients experienced an insufficient duration of effect with InA, and none with OnA. However, these lower rates may, in part, be due to variability in injection intervals in this sample, which could be because of scheduling considerations at the Harry S. Truman Veterans Health Administration Medical Center. In cases where OnA fails because of the development of antibodies, it might be reasonable to switch to InA treatment. Reformulation of InA with a pH-buffered solution may eliminate the difference in pain on injection. InA would then be a good alternative to OnA for treating CM.