The purposes of this longitudinal study were to examine techniques whereby kindergarten, Grade 1, and Grade 2 teachers might become more aware of differences in children which affect their performance and behavior, and to develop with teachers ways of accommodating classrooms to children and vice versa. Participants were 24 teachers from four school boards differing in locale served and in school sponsorship. Formal measures used included achievement tests, teacher ratings, classroom observations, and interviews. Data analysis revealed which teacher rating scales were strongly associated with tested academic skills, what combinations of teacher-rated characteristics were best associated with test scores, and relationships between "thrive" rankings and other variables. Additionally, 20 types of classroom problems were identified and categorized as involving constitution, non-social skills and behavior, self-direction and self-confidence, emotional functioning, and miscellaneous. Ratings did not predict specific problems, but only the incidence of problems in general. A total of 22 types of strategies were used by teachers to deal with problems they identified. Teachers' strategies were not associated with specific types of problems, suggesting that preventive techniques or "early identification" may be less useful than careful attention to individual children's current functioning and immediate situation. Appendixes, which take up more than half the document, contain statistical tables and tables giving percentages of straregy applications per problem by grades and by board. (RR)