Long term selection experiments with laboratory animals can provide an excellent opportunity to examine the assumptions underlying the predictions in animal breeding because of their short generation interval and high reproduction rate. The infinitesimal model of many genes each with small effect is often assumed for the analysis of economic traits in livestock species which have been selected by breeders for different objectives. Under such assumptions the genetic variance is expected to be reduced in a predictable way due to selection (Bulmer effect) and inbreeding in a selection experiment. The Animal Model with a complete relationship matrix back to the base population accounts for these reductions in genetic variance due to inbreeding and selection and thus provides an unbiased estimate of genetic variance in the base population (Sorensen and Kennedy, 1984). Lean mass in selected lines of mice was analysed using the individual animal model, to test if the infinitesimal model assumptions hold in the selected lines for this trait