Using Lehigh Longitudinal Study data (N = 457), the authors compare prospective parent self-reports and retrospective adolescent reports of early childhood physical abuse, exploring their correspondence, predictive equivalence, and outcomes associated with conflicting reports. Correspondence between prospective and retrospective reports of child maltreatment was moderate (φ = 0.27). Concurrence rates were similar for males and females. Analyses of the relative predictive capacity of prospective and retrospective measures revealed both to be significant predictors of key outcomes in adolescence. Findings support the predictive validity of both measures of childhood maltreatment and underscore the methodological challenges of measuring this important construct. Given the abundance and salience of research on the consequences of childhood maltreatment, greater attention to such measurement issues is due.