The objectives of this study were to estimate genetic parameters for milk performance traits of the East Friesian and Black-Brown milk sheep by using uni- and multivariate animal models. The analysis was based on 1746 lactation records including 9729 test day records of 1108 milk sheep distributed over 103 herds. The milk records were sampled from 1992 until 2000. Herds were located in Lower Saxony, Westphalia and Bavaria. The pedigree file contained 2988 animals. Using a multivariate test day model with fixed regression, the estimated heritabilities of the East Friesian milk sheep were for the milk, fat and protein yield h(2) = 0.17, 0.13 and 0.17, respectively. For the fat and protein content as well as the Somatic Cell Score (SCS) the heritabilities were h(2) = 0.07, 0.24 and 0.19, respectively. The heritability estimates of the milk traits for both breeds were similar to those of the East Friesian milk sheep, except for the heritability estimate of the protein content of the East Friesian milk sheep which was slightly higher. The additive genetic correlations between milk yield and fat as well as protein yield of East Friesian milk sheep were very high and positive (r(g) > 0.79), whereas the additive genetic correlations between milk yield and fat content were close to zero or even negative for the protein content. The estimated additive genetic correlations between protein yield and fat content were low and negative as well as the additive genetic correlation between fat yield and protein content. There were no considerable additive genetic correlations (r(g) = 0.14 to -0.17) between the SCS and the other milk performance traits. The genetic parameters estimated from field test records allow to achieve genetic progress in milk performance traits of milk sheep.
