We examined associations of social support and loneliness with eating and activity among parent–adolescent dyads (N = 2968) using actor–partner interdependence modeling. Loneliness had several actor associations with health behaviors (adolescents: less physical activity [PA], p < .001, more sedentariness, p < .001; parents: less fruit/vegetable consumption [FVC], p = .029, more hedonic food consumption [HFC], p = .002, and sedentariness, p < .001), but only one dyadic association (adolescent loneliness with less parent FVC, p = .039). Visible support was associated with less HFC, p < .001, and sedentariness, p < .001, but less FVC, p = .008, among adolescents. Invisible support was associated with less HFC, p = .003, but also less PA, p = .028, among adolescents. Both support types were associated with less HFC among parents, p < .001, but invisible support was also associated with less FVC, p = .029, and PA, p = .012, and more sedentariness, p = .013, among parents. When examining health behavior among parents and adolescents, it may be important to consider social support (but perhaps not loneliness) at a dyadic level.