The purposes of this study were to test the Bruhn and Parcel Model of Health Promotion (BPMHP) (1982), which explains how family influences and children's developmental characteristics interact to affect child health behaviors. The descriptive, correlational study was conducted using secondary data from the 117 adolescents included in Adolescents Benefit from Control of Diabetes Study. The results of structural equation modeling analysis indicated that greater parental modeling and positive parental interaction patterns were correlated with higher self-efficacy and lower levels of depressive symptoms among adolescents, which had a direct impact on better adherence. However, more parental modeling and interaction patterns did not have a separate direct effect on better adherence. Thus, the model generated by our data was significantly different from and so only partially supportive of the original BPMHP (1982). In addition, age and Tanner stages that can reflect changes in developmental characteristics were deleted from the concept called "developmental characteristics" in the BPMHP (1982), it may be that that concept should have been termed "psychosocial characteristics" rather than "developmental characteristics". In order to further explore the validity of the BPMHP (1982), future studies should be conducted to re-test this model.