Abstract Adolescent social anxiety (SA) assessments often include adolescent and parent reports, and low reporting correspondence results in uncertainties in clinical decision-making. Adolescents display SA within non-home contexts such as peer interactions. Yet, current methods for collecting peer reports raise confidentiality concerns, though adolescent SA assessments nonetheless would benefit from context-specific reports relevant to adolescent SA (i.e., interactions with unfamiliar peers). In a sample of 89 adolescents (30 Evaluation-Seeking; 59 Community Control), we collected SA reports from adolescents and their parents, and SA reports from unfamiliar peer confederates who interacted with adolescents during 20-minute mock social interactions. Adolescents and parents completed reports on trait measures of adolescent SA and related concerns (e.g., depressive symptoms), and adolescents completed self-reports of state arousal within mock social interactions. Adolescents’ SA reports correlated with reports on parallel measures from parents in the .30s and with peer confederates in the .40s to .50s, whereas reports from parent-confederate dyads correlated in the .07 to .22 range. Adolescent, parent, and peer confederate SA reports related to reports on trait measures of adolescent SA and depressive symptoms, and distinguished Evaluation-Seeking from Community Control Adolescents. Confederates’ SA reports incrementally predicted adolescents’ self-reported SA over and above parent reports, and vice versa, with combined R s ranging from .51 to .60. These combined R s approximate typical correspondence levels between informants who observe adolescents in the same context (e.g., mother-father). Adolescent and peer confederate (but not parent) SA reports predicted adolescents’ state arousal in social interactions. These findings have implications for clarifying patterns of reporting correspondence in clinical assessments of adolescent SA.