It seems safe to say that all official performance testing and progeny testing schemes for bacon pigs, up to the present, have been based on an implicit assumption that the pigs found to be superior under the management methods at the testing station would also be superior under other systems of management on commercial farms. If consideration is limited to reasonably good standards of management there is probably little danger in this assumption, though there is little direct evidence as yet to support it. The danger of mistakes in the selection of superior genotypes would be more likely to arise if there were systematic differences between the methods of management at the testing stations and on commercial farms. Actually such a difference does occur as regards methods of feeding. At testing stations, the pigs are generally fed more or less to appetite, being given as much as they will readily clear up in two feeds per day in England (N.P.B.A., 1957) and Northern Ireland (Heaney, 1962) and in three feeds per day in Denmark (Clausen & Thomsen, 1957). In the Republic of Ireland (Department of Agriculture, 1960) three feeds per day are given on a scale which must involve little, if any, restriction in the later stages of fattening. Such practices have the object of discovering the strains which become too fat before they reach bacon weights. Many farmers, on the other hand, adopt some form of restriction during the later stages of fattening, with the object of finishing the pigs at the right stage of fatness, and thus of improving the grading of the carcases at the factory.