Data were collected from one pedigreed generation of Norfa layers. 149 cocks and 477 hens were used. Each cock artificially inseminated three hens. 498 completely records of progeny were used. 15 selection indices were constructed by using different combinations of 4 traits and 3 sources of information. The studied traits were age at sexual maturity (SM), body weight (BW_M), egg weight (EW_M) and egg number (EN_42wk). The sources of information were own performance (OP), full sibs (FS) and half sibs (HS). The Least Squares and Maximum Likelihood General Purpose Program-Mixed Model "LSMLMW" (Harvey, 1990) was used to estimate heritability and phenotypic and genetic correlations of studied traits.       Heritability estimates of SM, BW_M, EW_M and EN_42wk were 0.32, 0.93, 0.17 and 0.57, respectively. The genetic and phenotypic correlations between SM, BW_M and EW_M were positive, while these correlations between each of these three traits and EN_42wk were negative. The value of each trait was affected by the source of information and number of traits in the index. By using the same source(s) of information, the value of each trait increased if the index was constructed by using 3 traits instead of 4 traits.       The accuracy of the index (r_TI) depends on number of traits; sources of information and the value of each trait were used to construct the index. There is a negative correlation between the value of omitted trait in the original index and the relative efficiency of the index. Excluding OP as a source of information from the index caused the highest decreasing in the accuracy of the index (r_TI). On the other hand excluding HS caused the lowest decreasing in the accuracy of the index. In relation to the original index (I₁), the most effective index (r_TI=100.6) was I₁₃ which include the 4 traits (SM, BW_M, EW_M and EN_42wk) and all available sources of information (OP, FS and HS). The main objective of this study was to evaluate a single source of information selection index versus multisource of information of information multi-traits selection indices in Norfa layers.