Objective: In the USA, few adolescents meet the national recommendations for fruit and vegetable consumption (FVC). This study examined the associations between two perceived food parenting practice scales (Promote and Ensure FVC Intake) and adolescent FVC, and assessed whether these associations were mediated by adolescent self-efficacy. Design: The study took the form of a secondary data analysis of publicly available data from the Family Life, Activity, Sun, Health, and Eating Study (FLASHE), a cross-sectional, Internet-based study administered by the US National Cancer Institute (NCI) between April and October 2014. Methods: Data from 1,354 parent–adolescent dyads were analysed. The PROCESS macro was used to conduct mediation analyses to estimate the indirect effect (IE) of each parenting practice on FVC through adolescent self-efficacy, adjusted for parental and adolescent characteristics. In addition, we examined whether adolescent age, gender, race/ethnicity or weight status moderated the mediation model. Results: The IEs were each statistically significant (Promote FV Intake: IE = 0.063, 95% bootstrap confidence interval [CI] = 0.046, 0.080; Ensure FV Intake: IE = 0.014, 95% bootstrap CI = 0.004, 0.024; p < .05 each), indicating significant mediation by adolescent self-efficacy. However, no adolescent characteristics moderated the effects of parenting practices on FVC through self-efficacy. Conclusions: Interventions should continue to promote parenting skills to improve adolescent self-efficacy and FVC. Insufficient evidence for moderation of the mediation effect by adolescent age, sex, race/ethnicity or weight status implies that the significant relationship between food parenting practices and FVC, mediated by self-efficacy, may operate similarly for adolescents that differ in each of these characteristics.