The use of structural equation modelling techniques for twin data are discussed. The technique not only allows estimation of genetic and environmental influences on traits or behaviours, it also gives information on how well such explanations of data fit genetic and environmental models of varying complexity. A model-fitting approach also allows the examination of more complex models than previous types of analysis of twin data provided. Applications to examine causal relationships between traits or behaviours, assortative mating, gene-environment interaction, multivariate analysis and longitudinal studies are briefly discussed. An example illustrating the use of model-fitting applied to twin data to estimate additive genetic, shared environmental and unique environmental factors is given.