Little hard evidence exists to support the contention that college remedial mathematics programs do indeed help students remove or remedy their deficiencies. The purpose of this study was to determine for remedial students the differential effects of two instructional modes centering around, respectively, (1) small‐group instructional processes (experimental), and (2) lecture‐demonstration processes (control). These effects were investigated in relation to (1) academic achievement, (2) attitude toward mathematics, and (3) sex difference. A total of 171 subjects participated in the main study. A 2 x 2 factorial analysis of covariance was employed to test the main hypotheses. Data from attitude and achievement scores were subjected to ranking and the Kendall tau (t) test of association was employed. The results of this study indicated the small‐group instructional procedures had a significantly greater positive impact (P<0‐001) on achievement in arithmetic, elementary algebra, and combined arithmetic and algebra than the lecture‐dmonstration method. Sex difference, in favor of females, was significant only in regard to algebra (P< 0‐002) and the combined content (P<0001). A general conclusion of this study is that it is feasible—given certain boundary conditions—to develop a remedial mathematics course whose main instructional focus is on students working in small groups.