Administrators spend most of their time on noninstructional matters. Because they excuse this by claiming that nothing is known of the learning process, they need a systematic ordering of the basic, verified knowledge on learning. For instance, findings in learning research show that the capacity to learn school subjects is far more widely distributed than was thought, that achievement is strongly interrelated with affective factors like self-esteem, and that testing can be used to improve teaching and learning processes as well as to classify students. Researchers have also gathered data on teaching activities and school structures and relationships that affect learning. Educational administrators can no longer maintain that we lack knowledge on learning processes, and they should take responsibility for using the knowledge gathered. However, most research and training in educational administration has focused on administrative behavior and matters unrelated to classroom learning rather than on school effects, student social patterns, and classroom-level variables. Among the implications of learning research findings are that administrators should reassert instructional leadership, foster staff development, and implement new school and classroom organizational patterns based on the research. (Author/RW)