Five years of research findings on assessment and decision making for learning disabled (LD) students are summarized through 14 generalizations. the generalizations deal with five issues: which students to refer for psychoeducational evaluation; which students to declare eligible for LD services; how to plan specific instructional interventions for individuals; how to evaluate the extent to which pupils are profiting from instruction; and how to evaluate the effectiveness of particular instructional programs. Data supporting each generalization are described briefly. Among implications are the needs to consider alternative approaches, to adopt an ecological model of students' learning and behavioral problems in educational settings, and to spend more time and resources in teaching and instructing rather than in testing and labeling. (Author/CL)