The ultimate breeding objective considered in this paper was to raise the profitability of laying duck producers by offering them a tool to increase body weight whenever accepting the expected consequences in terms of increase in weight of eggs and decrease in their number. This could be achieved by ameliorating the marketing body weight of surplus ducklings, normally used as a source of meat production, using information on their body weight and some body linear measurements recorded at early stages of life. Three selection indices were derived from data on 576 birds hatched and reared at the governmental, El-Serw Waterfowl Research Station, belonging to Animal Production Research Institute, located in Damietta, Egypt. These data represent records of 291 Domyati and 285 Khaki Campbell birds progeny of 40 sires and 160 dams. The traits considered as sources of information in the indices were body weight (BW), shank length (SL), keel length (KL) and breast girth (BG) recorded at 2, 4 and 8 weeks of age. In constructing the three indices (I w2 , I w4 and I w8 at weeks 2, 4 and 8, resp.) the aggregate genotype included body weights at 10 (BW w10) and 12 (BW w12) weeks of age. Phenotypic and genetic parameters were estimated from a multitrait-animal model where breed and sex effects were considered as fixed and the additive direct genetic effects as random. The results pinpointed that early selection hinged on I w4 (involving BW w4 , SL w4 , KL w4 and BG w4) as juxtaposed to I w8 (including BW w8 , SL w8 , KL w8 and BG w8) had comparable accuracy (0.47 vs. 0.49, resp.), expected gain in BW w10 (14.4 gm and vs. 14.7 gm, resp.) and expected gain in BW w12 (9.7 gm vs. 10.1 gm, resp.). Thereupon, I w4 would be the best possible index envisaging earliness, accuracy and efficiency.