Inasmuch as the newly established DSM-oriented CBCL/6-18 scales are to be increasingly employed to assess clinical/high-risk populations, it becomes important to explore their aetiology both within the normal- and the extreme range of variation in general population samples and to compare the results obtained in different age groups. We investigated by the Quantitative Maximum Likelihood, the De Fries-Fulker, and the Ordinal Maximum Likelihood methods the genetic and environmental influences upon the five DSM-oriented CBCL/6-18 scales in 796 twins aged 8-17 years belonging to the general population-based Italian Twin Registry. When children were analysed together regardless of age, most best-fitting solutions yielded genetic and non-shared environmental factors as the sole influences for DSM-oriented CBCL/6-18 behaviours, both for the normal and the extreme variations. When analyses were conducted separately for two age groups, shared environmental influences emerged consistently for Affective and Anxiety Problems in children aged 8-11. Oppositional-Defiant, Attention Deficit/Hyperactivity, and Conduct Problems appeared-with few exceptions-influenced only by genetic and non-shared environmental factors in both age groups, according to all three computational approaches. The De Fries-Fulker method appeared to be more sensitive in detecting shared environmental effects. Analysing the same set of data with different analytic approaches leads to better-balanced views on the aetiology of psychopathological behaviours in the developmental years.