A study examined the effect of Direct Instruction (DI) on the reading achievement of sixth graders. Subjects were 72 Chicago Public School students, nearly 100% minority. In the school year of 1994-95 no students received DI. A sample of 30 students who performed poorly, with stanines of 1, 2, and 3 on the Iowa Tests of Basic Skills (ITBS) were chosen to receive reading instruction through the DI method in 1995-96. The remaining students in the grade level did not receive the program. Results of the 1994 ITBS were used as a pretest and the 1995 scores were used as a posttest to measure students' gains within the year. Results indicated that the experimental group's gain from 1994-95 was 1.06, and the control group's was .45. These findings suggest that the Direct Instruction model should continue. If the experimental group did not receive DI and had experienced a year's growth of only .45, they would be even farther behind than in the previous year. Further research using a larger sample population should be undertaken. Other recommendations are that teachers of DI should receive continual inservice training and that class size for students receiving Direct Instruction should be reduced. (Contains 2 tables of data and 10 references.) (CR)