We investigated whether selective discussion of autobiographical memory narratives would impact the quality of young people's recall of their nondiscussed memory narratives. Children (ages 8-9 years, n = 65) and adolescents (ages 13-15 years, n = 58) completed an adapted version of the retrieval-induced forgetting (RIF) paradigm for self-generated positively and negatively valenced autobiographical memories. Overall, 54.5% of the sample were female and 63.4% were of European ethnicity (11.4% Pacific Peoples, 8.1% Middle Eastern/Latin American/African, 7.3% Maori, 7.3% Asian, 2.4% Other). We defined narrative qualities as narrative coherence (Reese et al., 2011) and episodic and nonepisodic information (Addis et al., 2008). In light of developmental findings in other domains of autobiographical memory research (e.g., Reese et al., 2011; Willoughby et al., 2012), we hypothesized that selective discussion would result in RIF for children's, but not adolescents', narrative coherence and episodic detail, and that RIF would not occur for nonepisodic details for either children or adolescents. Findings for narrative coherence and nonepisodic detail indicated support for our hypotheses. Findings for episodic detail were in partial support of our hypothesis; RIF for episodic detail was found for both children and adolescents. Our findings not only demonstrate the importance of investigating the wider effects of RIF but also uncovered developmental differences previously overlooked in the field.