Abstract: The records and responses to selection were analyzed in 4 generations of three pure and commercial lines of 31, 103 and 107 Iranian silkworm. In the beginning year of experiment, a parental group was selected, recorded and coded for each 3 lines of 31, 103 and 107 which had further weight average than the population average. These parents were crossed with each other and a selected group was made for these three lines. Also a parental group was chosen, recorded and coded for each of the three lines which include the weight average equal to population average weight. These parents also were crossed with each other and a coincidental group was made for these three lines. The members of each group were grown up and crossed during three successive generations coincidentally with out any re selection. The 27 average records were registered and analyzed. The data analysis in this research showed that among three studied lines in the whole year for larva vitality percentage, the highest response to selection is shown by line 103 (28.62%) then line 31 (70%) and their least line 107 (-3.90%) in the base population. The comparison of average response to these three lines selection also shows that the difference of response to the selection between lines is not significant statistically (p>0.05). Among lines in the whole year for larva remaining percentage, the highest response to selection was observed in line 31 (33.13 cocoon) followed by line 103 (17.91 cocoon) in the base population. Also, the least response to the selection in base population belonged to line 107 (9.29 cocoon). The comparison of average response to these lines selection showed that the response difference to selection in lines has not been significant statistically (p>0.05). Among the three studied lines in a whole year for a best cocoon total weight, the highest response to the selection in base population belonged to line 31 (49.44 g) then in line 103 (14.02 g) and the least response to selection in the base population belonged to line 107 (10.90 g). The comparison of average response to these three lines selection showed that there was not any statistical meaning in response difference to the selection among them (p>0.05). Among lines in the whole year for best single cocoons weight, the highest response to the selection in base population belonged to line 31 (0.05763 g) and then line 107 (0.05142 g) and the least response to selection belonged to the line 103 (0.014.7 g) in the base population. The comparison of average response to selection between three lines has shown that the response difference to the selection among lines is not significant statistically (p>0.05). Through lines in a whole year for the cocoon weight resulted form 10000 larva, the highest response to selection in the base population belonged to line 31 (601.1 g) then line 107 (572.9 g). The least response to the selection was observed in line 103 (-41.1 g). The comparison of average response to the selection through lines show that the response difference to the selection among these three lines is not significant statistically (p>0.05).