This systematic review found that studies of adverse drug event rates in ambulatory care yielded widely differing estimates of error prevalence depending on the error detection method used and the specific population studied.