The current investigation assessed the effectiveness of protective equipment, specifically arm splints, in reducing the occurrence of severe self-injurious behavior (SIB). Although the protective equipment reduced rates of SIB to near-zero levels, self-restraint subsequently emerged. In an attempt to reduce self-restraint while maintaining reductions in SIB, we provided noncontingent access to preferred stimuli. The presentation of preferred stimuli along with the use of protective equipment reduced both the rate of SIB and the duration of self-restraint. (Contains 1 figure.)