Two self-contained fourth grade classrooms were compared during a 1-year study conducted in a small rural community in Indiana. Pre-test measures consisted of the previous year's third grade scores on the Cognitive Abilities test and a self-developed inventory of attitude toward school and computers. The control group (n=28) received instruction throughout the study via traditional teaching methods with one computer available to the class. The experimental group (n=29), with a student/computer ratio of 2:1, began the year by learning keyboarding and becoming familiar with the computer system and the available variety of software, as well as attending to the traditional lessons with their teacher. After this initial instruction, these students spent a minimum of two hours per day at the computer, either alone, or with a partner, working with software from all areas of typical daily instruction. The measures of posttest performance for the two groups compared computer skills mastered; problem solving ability; and the Iowa Test of Basic Skills Reading, Math, and Composite Subtests. Both the raw scores on these measures and the scores adjusted for differences in intelligence were compared. Student attitudes toward school and computers, and teacher perceptions of student abilities were also compared. Although the experimental group had significantly higher scores on the computer skills test, none of the other measures produced significant results. Five appendixes and two supplemental analyses provide the study data and statistical analyses, sample measuring instruments, and a list of computer skills objectives. Ten references are included. (EW)