Abstract A combination of twin and parent-offspring data on the EPQ and JEPQ is subjected to genotype-environmental analysis by the balanced pedigree method. The method, based upon Jöreskog's approach to the analysis of covariance structures, employs the principle of maximum-likelihood and permits tests of the model and the extraction of standard errors for parameter estimates. The data for extraversion and neuroticism are on the whole consistent with a simple model which assumes additive gene action, random mating and environmental effects within families. For the psychoticism dimension and the lie scale mating is found to be non-random. The stability of personality expression across generations is investigated. With the exception of the neuroticism dimension there is a marked inconsistency of gene action between juveniles and adults. For the lie scale social interaction between juvenile cotwins is detected and the juvenile scale is suggested as a paradigm of a trait for which the environmental interactions between relatives have a major role in the causes of individual variability.