This study investigated curriculum influences on student mathematics achievement by following two groups of students from fifth to sixth grade that were taught either the reformed curriculum or the conventional curriculum. Analyses with three-level modeling were conducted to examine learning outcomes of the students who were assessed three times over a period of 18 months. Achievement was measured with regard to computation, routine problem solving, and complex problem solving. Affective aspects included self-reported interest in learning mathematics, classroom participation, views of the nature of mathematics, and views of learning mathematics. The results showed overall improved performance among all the students over the time on computation, routine problem solving, and complex problem solving but not on the affective measures. There were differentiated patterns of performance between the groups. On the initial assessment, the reform group performed better than the non-reform group on calculation, complex problem solving, and indicated higher interest in learning mathematics. The two groups did not differ on the other achievement and affective measures at the first time of assessment. There was no significant difference in growth rate between the groups on the cognitive and affective measures except that the non-reform group progressed at a faster pace on calculation. Therefore, the non-reform group outperformed the reform group on computation at the third (last) assessment. These results are discussed with respect to the possible influence of the curriculum on student learning. (Contains 4 tables and 4 figures.)