For the UK to continue as the major producer and exporter of lamb in Europe, steps must be taken to improve carcass quality in an effort to safeguard existing markets and develop new ones. Each stratum of the sheep industry has to address this goal. Carcass quality can be changed by selection, and this is now being taken advantage of in terminal sire breeds and, to a lesser extent, in hill breeds. However, there has been little progress in the longwool breeds used in the production of halfbred ewes. The objectives of this project are therefore to provide essential information for the development of a multi-trait selection index for improving carcass quality of longwool sheep and their halfbred progeny without compromising reproductive performance and maternal ability, using the Bluefaced Leicester (BFL) as a model. The project has been designed in a stepwise fashion with each objective building upon another. During Phase 1, BFL rams of known genetic merit will be selected from a relatively large pool created by genetically linking flocks together through a Sire Reference Scheme. These BFL rams will be mated to two genotypes of hill ewes, and the halfbred (Mule) wether lambs will be assessed for growth and carcass traits. In addition, computer modelling will be used to identify optimal breeding structures for the production and utilisation of halfbred ewes. During Phase 2, reproductive and other maternal traits will be assessed in Mule ewes. Finally, using the information gathered, genetic and phenotypic correlations will be estimated using appropriate statistical techniques, and a multi-trait selection index will be constructed for a defined set of traits appropriate to longwool sheep. The project encompasses a national perspective with experimental sites in England, Wales and Scotland which, together with a Steering Group comprising key sheep industry representatives and a comprehensive promotional programme, will ensure maximum impact on, and an effective transfer of technology to, the sheep industry as a whole.