On one hand, the mediational sequence from body dissatisfaction through dieting to bulimia — often referred to as the "restraint pathway" — has been empirically validated across numerous methodologies and samples of adolescent girls; on the other hand, the prevalence rate of bulimic pathology in adolescent girls pales in comparison to rates of body dissatisfaction and dieting. This discrepancy indicates that the restraint pathway may only be applicable to individuals possessing certain characteristics or experiencing certain circumstances, and represents the crux of the current study. Given that higher levels of thin-ideal internalization and interoceptive deficits have distinguished symptomatic from asymptomatic youth in other research (e.g., Lowe et al., 1996; Mintz & Betz, 1988), it is reasonable to examine their moderating potential in the context of the restraint pathway. Also, because eating disorder risk increases with age throughout adolescence (e.g., Jones, Bennett, Olmsted, Lawson, & Rodin, 2001), it is worth exploring whether the obtained pattern of relations differs in strength for early-, mid-, and late-adolescent girls. Using self-report data from a nonclinical, cross-sectional sample of 353 middle-school (n = 115; Mage = 12.77), high-school (n = 112; Mage = 15.95), and college girls (n = 126; Mage = 19.39) and a moderated mediation approach (see Hayes, 2013, pp. 393-397), we found that: (1) thin-ideal internalization moderated the relation between body dissatisfaction and dieting, such that body-dissatisfied adolescent girls who reported a high, versus low, endorsement of sociocultural thinness norms engaged in greater dietary restraint; (2) interoceptive deficits and age multiplicatively moderated the relation between dieting and bulimia, such that only dieters who reported difficulty perceiving internal body cues and were of college age expressed bulimic tendencies; and (3) the indirect effect of body dissatisfaction on bulimic symptoms through dieting pertained only to college girls who reported high interoceptive deficits, but this effect was significantly stronger for those who reported high, versus low, thin-ideal internalization. Our results simultaneously contribute to (i.e., for late-adolescent girls) and undermine (i.e., for early- and mid-adolescent girls) the specificity of the restraint pathway— a paradox around which theoretical, empirical, and practical implications are discussed.