Data from the National Longitudinal Study of Adolescent Health (Add Health) provided the study sample of 5,201 adolescents, with 1,750 adolescents reporting loneliness. Logistic regression procedure was used to analyze the main effects and interaction terms of the study variables. Lonely adolescents were found to be older and female. Adolescents from one-parent and blended families were more likely to be lonely when compared to adolescents from intact families. Adolescents who felt less support from their parents, had fewer opportunities to be with parents throughout the day, and participated in fewer activities with parents were more likely to be lonely. When the interaction between parent-child relationship factors and family type was considered in the logistic regression model, the risk for loneliness was reduced. The study findings emphasized the importance of the continuing need for parental attention, especially in situations of changed family structure. Along with gender and age, parent-child relationship factors and family structure were important predictors of adolescent loneliness. The effect of family structure can be modified by parent-child relationship factors in reducing the risk of adolescent loneliness. 5 tables and 35 references