The effectiveness of using computer courseware developed for adults reading below the fourth-grade level was investigated with inmates at Rockview State Correctional Institution. Data were collected both to establish multiple baseline data and to permit experimental group comparison with a control group. The multiple baselines portion of the study involved 11 male inmates with reading levels of 0-4 who completed 18-25 hours of work with the courseware. The Wide Range Achievement Test (WRAT) and the Slosson Oral Reading Test (SORT) were used as pre- and posttests. The courseware was considered effective. Testing 10 months later showed that students' retention level was about 90 percent. In the experimental/control group portion of the study, 27 inmates were tested with 2 sections of the Baltimore County Design, the Bader Reading and Language Inventory, the SORT, and a test of incidental learning of computer words. The experimental group each used the courseware for 10-22 hours whereas the control group received small group and tutorial instruction using traditional beginning materials for adult reading instruction but was not exposed to computer-assisted instruction. The 23 men who completed the work were given posttests using the same measures used for the pretests. Analysis of variance for repeated measures was used to analyze the test differences. The only measure that showed a statistically significant difference was the Bader Reading and Language Inventory. The control group showed no improvement on the posttest measure and had fallen behind the experimental group at that time, even though they had started out at a higher level. (Two references and three tables are included.) (CML)