The two studies established an empirically derived set of guidelines for producing and evaluating microcomputer courseware for students with learning disabilities. About 100 teachers of learning disabled students in Pennsylvania and Virginia participated in field testing the guidelines. The first study was conducted in four steps: first, development and validation of the instrument; second, establishment of interrater agreement on one piece of courseware; third, rank ordering by teachers of the importance of criteria on the instrument; and fourth, use of the instrument to rate four pieces of courseware. Study 2 duplicated the third step of the original study with different teachers. The conclusions of the study established nine essential criteria for evaluating and producing software: (1) directions the students must read should be simple; (2) alternate means of presenting the same concept must be provided; (3) the screen must be uncluttered; (4) minimal keyboard skills should be necessary; (5) feedback regarding correctness or incorrectness must be provided; (6) adequate opportunity to review concepts must be provided; (7) very basic skills that non-handicapped students would learn incidentally must be taught; (8) prompts to keep students moving through the program must be provided; and (9) the courseware must aim for mastery learning. The questionnaire and tables are appended. (Author/DB)