|Abstract||Mass selection of individuals in seedling or early clonal stage trials is routinely used in most sugarcane improvement programs throughout the world. It is, however, inefficient as the Heritability of cane yield on a single plant basis is low. In Australia, the introduction of mobile truck-mounted weighing equipment offered the opportunity to implement family selection utilising weighed family data. Family selection has been used in some Bureau of Sugar Experiment Stations' (BSES) selection programs since 1986 and is now routinely used in all regional selection programs. This research has demonstrated that a combination of family and mass selection in stage 2 of selection will result in larger genetic gains and a higher frequency of superior clones in later stages than mass or family selection alone. This combination allows improved efficiency since fewer resources are required to select only within superior families in the first ratoon crop than to mass select individuals across the entire population. A liberal family selection rate (about 40%) balances genetic gain and the need to maintain a broad genetic base. While mass selection rates did not vary in this work, it is believed that differential seelction rates within families sould be used so that more clones are selected out of the best families. The availability of objective family data also allows more accurate estimation of the breeding value of parents utilising best linear unbiased predictors (BLUP). This results in better genetic combinations through crossing and provides more objective information on new parents.