While diagnosis is generally considered a vital element in reading clinicians' expertise, research has revealed that even degreed, experienced reading clinicians display little personal consistency or agreement with one another when diagnosing simulated cases of reading difficulty. Three studies were conducted to determine if systematizing the diagnostic process by providing a process model, diagnostic decision aids, and sufficient practice with feedback would result in more reliable diagnoses. Subjects were (1) master's degree students in reading who had some prior course work in diagnosis, (2) master's degree candidates with prior teaching experience and coursework in reading, and (3) experienced classroom teachers with little or no formal training in reading or reading diagnosis. The results indicated that the training was successful, both with degreed reading clinicians and with teachers who had no previous work in reading diagnosis. (Appendixes contain the cue inventory for one simulated case, a portion of a diagnostic decision aid, a portion of a diagnostic checklist, and tables of data from the studies.) (Author/FL)