Injuries are a major public health problem among the elderly, and they have substantial medical, social, and economic consequences. There has been a longstanding concern that psychotropic drugs increase the risk of injury, particularly in the elderly, who frequently use psychotropic drugs and are more vulnerable to drug effects that may predispose to injuries. For benzodiazepines and antidepressants, the most commonly used psychotropic drugs outside of institutional settings, an extensive body of data demonstrating psychomotor function impairment establishes a firm basis for concern, which is reinforced by the more limited epidemiologic data. However, because of limitations in both types of data, this issue remains controversial. To resolve this controversy, we suggest three lines of further research: psychomotor function tests more directly relevant to the injury risk of the elderly medication user, epidemiologic studies designed to distinguish drug effects from those of underlying psychopathology, and clinical trials that include injuries as an outcome.