This study was carried out to expect genetic gain and possibility of genetic improvement of pre-weaning growth traits of Friesian heifers. Records of 1748 progeny of 61 sires and 846 dams were collected from 1995 to 2010. That belongs to the Experimental and Researches Unit of Animal Production in Tokh Tanbisha, in the middle Delta. This belongs to Faculty of Agric., Minoufiya Univ., Egypt. Analysis was carried out by SAS (SAS, 2002) to test the significance of factors affecting studied traits and MTDFREML program (Boldman et al., 1995) with the mixed model used including sires, dams as random effects and parity, year and season of birth as fixed effects to estimate the variance components. Furthermore, Selection Index Program (Wagenaar et al., 1995) and Matlab program (Matlab, 2002) were used to construct the selection indices. It is using multi-source multi-trait selection index methods which include general, restricted indices at different levels of restrictions on increase in birth weight, multi-source index and selection index with desired genetic gain. Estimates of variance components for body weight at birth, 30, 60 and 90 days of age were computed and used to construct selection indices to improve studied traits in Friesian heifers. The general index incorporating weights of birth _(W0), 30 _(W30), 60 _(W60) and 90 _(ww) days of age had the highest correlation with aggregate breeding value (R_ih=0.5051). The correlation fell to 0.321 when body weight at birth was restricted by 100% from the index. Furthermore, Using multi-source of information will enhance correlation with aggregate breeding value (R_ih= 0.5208) and the expected genetic gain for all studied traits and will be useful in case of missing data for selected animals and/or weight at 90 day. The highest expected genetic gain in W90 was 1.021 kg at (i=1) and 0.3574 kg at (i=0.35) per generation with multi-source index (I_6-multi); this decreased to 0.740 kg/generation when body weight at birth with full restricted strategy (I_{5 (100%)}).  From the present results, it could be using (I_6-multi,, I₁,and I_8-desired) to improve pre-weaning body weights in Friesian heifers and using (I_{4 (75%}) under restriction strategy in case of population that have already not reached optimal W₀ but in case of population that have already reached optimal W₀ using completely restriction index (I_{5 (100%)}).