A multidimensional peer nomination inventory (MPNI; 30 items) and parallel versions--MPNI-Teacher and MPNI-Parental Rating Forms (37 items each)--were developed during a major new Finnish study of families with twins. The twins (477 girls and 467 boys) were 12 years old, representing subsets of three nationwide Finnish twin cohorts (b. 1983-1985). They were enrolled in 503 school classes, and the total number of children participating in peer nominations was 12,937. Three main factors were extracted from peer nominations and teacher and parental assessments. Intercorrelating sub-components were found, especially in parental assessments. Scales were formed, accordingly, for Behavioral Problems (including Hyperactivity-Impulsivity, Aggression, and Inattention), Emotional Problems (including Depression and Social Anxiety), and Adjustment (including Constructiveness, Compliance, and Social Activity). A framework for the development of the multidimensional inventory was a model of emotional and behavioural regulation. Peer nominations were most reliable, while parental assessments, although mostly satisfactory, were least reliable. Results provided evidence of concurrent validity of peer-referenced assessment, using teacher assessments as criteria; correlations between assessments of peers and parents were lower. The inventory has discriminative validity. Intra-pair correlations of monozygotic co-twins were higher than correlations of same-sex (SS) and opposite-sex (OS) dizygotic (DZ) co-twins for all scales across all three assessors, and peer nominations of both SS and OSDZ co-twins yielded correlations significantly greater than zero for all scales. All scales, except Depression and Social Anxiety, differentiated boys from girls.