A research paradigm, including an AB1AB2 single-subject design was developed and used in conjunction with an adult-child interaction analysis methodology to evaluate the effectiveness of a behavioral treatment approach to a feeding problem in a 2 1/2-year-old male. The results indicate that after nine treatment sessions, the child's diet was appropriate and the undesirable mother-child interaction pattern was changed. In addition, the interaction analysis methodology was used to evaluate the therapist's behavior modification techniques so that these could be taught to the mother. This study continued the previous endeavors to delineate feeding problems in children and subsequently to treat these problems with a behavioral approach. This kind of systematic research, with functional analysis of presenting problems and specification of the therapist's treatment behavior, is offered as a clinically appropriate paradigm for psychotherapy and behavior modification research.