The major purpose of this study is to investigate the contribution of homework andrnother va riables to students' mathematics achievement. The subjects were 240rnrandomly selected yth graders (n =1 20) and 8th graders (n=1 20) from Awassa TaborrnElementary School in 1996 Eth.C. Each grade students were assigned intornexperimental and control groups, and the former were taught with assignment ofrnhomework and the latter learned without assignment of homework for eight weeks.rnAchievement tests and questionnai re were instruments for data collection. Initia lly,rnthese instruments were administered on a pilot sample and improved through itemrnanalysis. The data analysis was carried out using ANOVA, F-test, t-test, the TukeyrnMethod and omega squared (w2rn) . Results ind icated that there is significantrndifference in mathematics achievement between students in experimental andrncontrol groups within achievement levels of each grade. Specificall y, the effect ofrnhomework accounted 19.2% , 17.2% and 22.6% of the va riances in mathematicsrnachievement of yth grade high achievers, average achievers and low achievers,rnrespectively. With rega rd to grade 8 students, its effect accounted 21.1 % of thernvariance in high achievers' mathematics achievement, 19.7% of the variance inrnaverage achievers' mathematics achievement, and there was no significantrndifference observed between low achievers in homework and no homework groupsrnat alpha 0.05 level. The results also revealed that there wasn't stati sticallyrnsign ificant difference in attitudes toward homework between students in bothrngroups within achievement levels of each grade during the cou rse of the study