Research related to the use of mechanical restraint devices (e.g., arm splints) has been limited despite the frequency with which they are used in the clinical treatment of severe and dangerous self‐injurious behavior (SIB). Wallace, Iwata, Zhou, and Goff (1999) used a rapid restraint analysis (RRA) to evaluate the effects of different levels of arm‐splint flexion on rates of hand‐to‐head SIB and adaptive behavior. The goal of the current study was 3‐fold: (a) to extend previous research on RRA by reporting RRA results for 10 participants and investigating post‐RRA outcomes, (b) to investigate whether the RRA might have the potential to eliminate the need for restraint fading altogether for some individuals, and (c) to investigate whether the RRA might help to identify a starting point for restraint fading for individuals for whom fading is necessary. Results suggested that the RRA helped to eliminate the need for restraint fading for 6 participants. With respect to the identification of a starting point for restraint fading, results were mixed. Overall, 8 of the 10 participants were discharged from the inpatient unit at a less restrictive level of flexion or at the same level of flexion identified by the RRA. Implications for adoption of the RRA are discussed.