In the present study we used an empirical validation approach to determine psychosocial variables related to adaptive management of diabetes in two samples of diabetic children and their families. Children from the Kansas University Medical Center (n = 31) and children from the Virginia Diabetes Association (n = 25) provided results of the glycosylated hemoglobin test (a measure of diabetic control); completed self-report measures on variables of anxiety, coping, family adjustment, and health locus of control; and underwent structured interviewing. Parents completed parallel self-report measures, as well as an assessment of child behavior. In addition, a subsample of parent-child dyads participated in a role-play measure of family interactions around the child’s diabetes. All demographic variables and psychosocial measures were not Significantly different across the Kansas and Virginia samples. Children in optimal control of diabetes had more structured and organized family environments, more frequently believed that ''powerful others'' were responsible for their health, and showed lower levels of negative coping with family problems than children in nonoptimal diabetic control. Surprisingly, children in optimal diabetic control showed poorer knowledge of diabetes and its management than children in nonoptimal control. Psychological interventions to promote adaptive diabetic control should incorporate findings related to family environment, locus of control, and the child’s style of coping with family problems.