This study is an original investigation of the convergent validity of the Friends and Family Interview (FFI) with the Inventory of Parent and Peer Attachment (IPPA), involving teenagers living in either low-risk or at-risk contexts. Participants were 130 adolescents (12–19 years), 91 low-risk (community) and 39 at-risk (in residential-care). Aims were to explore: (1) The convergence between the FFI and the IPPA in the two samples; (2) The predictive role of FFI insecure attachment scales on IPPA’s responses. Results revealed that FFI and IPPA responses were shown to significantly converge among the low-risk, mainly securely attached, community teenagers, but not among the youth living in high-risk circumstances where most interview responses were insecure. Lower convergency in the at-risk sample is likely due to idealizing or normalizing response to the questionnaire provoked by a defensive need to downplay or otherwise ignore adverse experiences, a phenomenon detectable only through careful coding of the language in interview responses that reveal the high levels of idealization or anger typifying these insecure (dismissing or preoccupied) interviews. Therefore, mixed methods are advisable, particularly with at-risk teenagers.