This study compares the cost-effectiveness of an individualized basic mathematics program with a traditional program at the junior high school level. Eighth-grade basic mathematics students (N=118) were identified in two schools. The racial and ethnic makeup, socioeconomic level of residents and school size differed between schools. Teachers were assigned an experimental and control group matched for mathematical abilities. The preassembled groups were assigned to a nonrandomized control-group pretest-posttest design. Achievement gains were measured by the Comprehensive Test of Basic Skills (CTBS). Experimental and control groups' average of months gained in achievement was divided by the cost per pupil to derive a cost-effectiveness factor. The effects of the treatment, the achievement scores in different schools, and the interaction between the treatment and schools were analyzed by four two-way analyses of variance. The achievement gains and the cost of the programs were dependent variables; the treatment received and school were the independent variables. The operational cost-effectiveness factor was $0.54 per unit achievement gain for the experimental program and $0.36 for the traditional. The experimental groups had significantly high arithmetical applications scores, and approached significance on the total mathematics scores. (Author/JP)