The effectiveness of a diagnostic teaching program in teaching reading to emotionally disturbed and socially maladjusted boys in two schools in New York City was investigated. The diagnostic teaching program specified behavioral objectives which fit the individual student, prescribed instructional styles for the teacher, and outlined techniques of diagnosing individual learning needs. It was hypothesized that the experimental group of about 40 boys, aged 9 to 14 years, would score significantly higher on reading tests than the matched control group. This hypothesis was not supported by the data from the post-tests, although the experimental group did show significantly higher subtest scores in letter names, short sound vowel recognition, short sound vowel production, and two-letter consonant blends. Analysis of correlations of learning mode and score gains suggested that more gains were achieved (1) when motor activities were used as a learning mode, (2) when reading material was relevant to the student's experiences, (3) when differing experiential and knowledge levels were accounted for in instructional materials, and (4) when teachers used a variety of equivalent reading strategies to compensate for short attention span. Tables, a bibliography, and sample diagnostic tests are included. (AL)