The present experiment was planned to study the effect of selection for higher three week body weight on overall growth performance in Japanese quail for three generations. Initially a total of 11000 quail chicks were procured from the hatchery of ART Centre and randomly divided into 22 groups with equal number of chicks in each. At the age of 21 days birds were weighed individually and the birds (male and female) with the highest body weight were selected to be the parents of next generation. Out of these 22 groups, 20 groups were subjected to mass selection. In one group (group #21) selection was performed with full pedigree records. In group 22nd, birds were picked up randomly to be the parents of next generation without performing selection. The same procedure was repeated for all the three generations. The data on each flock in each generation were recorded regarding different parameters directly or indirectly linked with growth performance. The data were subjected to Analysis of Variance (ANOVA) in Completely Randomized Design (CRD) under factorial arrangements using GLM (General Linear Model) procedures with the help of SAS, 9.1. and the comparison of means using Duncan’s Multiple Range (DMR) test depicted significant improvement in feed intake (g), body weight (g), body weight gain (g), FCR, Caloric and Protein intake / g body weight gain and decreased mortality in groups subjected to selection either mass or pedigree based while leg abnormality revealed non-significant differences in all the three groups.