Although some aspects of the educational effects of computer assisted instruction (CAI) are still unclear, some research findings are now quite definite. Two meta-analysis reviews of CAI have been completed at the University of Michigan, one on secondary school teaching and one on college teaching, and their findings have been independently verified by independent meta-analyses. Conclusions are that CAI (1) has real potential as a tool in improving student achievement in precollege classes; (2) fosters positive attitudes toward the computer; and (3) can produce substantial savings in instructional time. Findings on CAI outcomes are very robust, with differences in instructional environments, samples of students used, and methodological designs of studies having no substantial effects on outcomes of CAI and conventional teaching. The effectiveness of CAI may be limited by its use as a complete replacement for conventional instruction and enhanced by the use of up-to-date programs and computers. A safe conclusion is that the computer can be used to help learners become better readers, calculators, writers, and problem solvers. Six references are listed. (LMM)