Behavior-analytic approaches to occupational safety are often effective for improving safety in organizations, and have been successful in a wide variety of settings. The effects of these safety processes are thought to arise primarily from the behavioral observation process and the delivery of feedback. Typically, supervisors or employee observers involved in behavioral safety implementations conduct observations. The present study was an attempt to assess the effects of conducting observations on an observer's safety performance. An ABC multiple baseline counterbalanced across two sets of behaviors was conducted in a simulated office. The target behaviors involved knee and back positions during lifts; back, shoulder, and feet positions while sitting; neck and wrist positions while typing; and neck position during phone use. Substantial improvements in safety performance occurred after participants conducted observations on a videotape of a confederate's performance. The possible behavioral functions responsible for this change, and the implications of these findings for applied settings, are discussed.