in animal breeding applications, more than one traits are measured on each individual, for example, beside milk production the amount of fat and protein in milk are also measured. Since these traits are strongly correlated, instead of analysing each one separately, it is much more appropriate to evaluate them simultaneously. Thus estimates of genetic and phenotypic correlations and heritability are easily obtained. The obvious problem of multiple-trait analysis is the additional computational requirements due to the increased number of equations to be solved. The other problem is that the more traits are analysed, the greater the probability of estimates outside the parameter space.