Adult observers are typically the only informants on emotional and behavioral problems in young children. Although additional information can be provided by child self-report, few validated, structured instruments are available to obtain self-report from young children. The Berkeley Puppet Interview (BPI) has been developed to obtain structured self-reports on multiple domains of mental health and social well-being. This study was the 1st to evaluate the psychometric properties of the BPI in a large sample. We studied 8 a priori scales of the interview in a Dutch community sample of 6,375 children ages 5-7 years. Using confirmatory factor analysis, we demonstrated adequate fit (Tucker-Lewis index = .90; comparative fit index = .90; root-mean-square error of approximation = .03) of a multidimensional model with 50 items loading on 8 latent factors (Depression, Separation Anxiety, Overanxious, Oppositional Defiant, Overt Hostility, Conduct Problems, Bullied by Peers, and Peer Acceptance/Rejection). This model was invariant across gender. Children reported anxiety-related problems more frequently than depressive problems, behavioral problems, or difficulties in peer relations. Reliability analyses showed that 3 broadband scales designated as Internalizing, Externalizing, and Peer Relations were homogeneous constructs (alpha s = .68-.79). Higher scores on most BPI scales were associated with lower maternal education, lower family income, and non-Western ethnicity. Boys reported more behavioral and peer relation problems, whereas girls reported more emotional problems. The findings indicate that young children from socioeconomically and demographically diverse backgrounds are capable of providing valid, multidimensional information on their emotional, behavioral, and peer relation problems using the BPI. Young children's self-report is a promising addition to existing assessment tools.