Behavioral assessment procedures were used to determine the maintaining conditions of self-injury exhibited by 2 children with severe multiple handicaps. For both children, negative reinforcement (escape from grooming activities) was determined to be the maintaining reinforcer for self-injury (hand/arm biting) within an alternating treatments design. The treatment packages involved the use of negative reinforcement (brief escape from grooming activities) contingent upon a behavior that was incompatible with self-injury (reaching and pressing a microswitch that activated a prerecorded message of "stop"). Treatment was evaluated with a reversal design for 1 child and with a multiple baseline across grooming activities for the 2nd child. The treatment led to a marked decrease in self-injury for both children. At follow-up, high rates of self-injury were reported for the 1st child, but low rates of self-injury and an increase in task-related appropriate behavior were observed for the 2nd child.DESCRIPTORS: functional analysis, negative reinforcement, severely handicapped, self-inju- lection of an appropriate treatment for SIB was dependent on the maintaining conditions.Examples of the use of hypothesis testing or functional analysis procedures to prescribe treatments were reported by Repp, Felce, and Barton (1988) and Steege, Wacker, Berg, Cigrand, and Cooper (1989). Repp et al. (1988) successfully treated self-injurious and stereotypic behaviors of 3 severely handicapped children by first conducting behavioral assessments of self-injurious or stereotypic behavior during baseline conditions and then forming hypotheses about the variables maintaining the maladaptive behaviors and matching individual treatments. The results showed that a successful treatment program can be developed based on hypotheses regarding why a behavior occurred during baseline. Steege et al. (1989) (wiNnR 1990)