As part of the Competency Based Teacher Certification Project in Carroll County, Georgia, large samples of elementary and secondary school teachers and students were observed during a two-year period. Four low-inference observation measures were used to record teacher behaviors and student-teacher interactions: (1) Teacher Practices Observation Record (TPOR); (2) Spaulding Teacher Activity Rating Schedule (STARS); (3) Observation Schedule and Record (OScAR); and (4) Florida Classroom Climate and Control System (FLACCS). Three major areas of student growth were also examined: cognitive gains on standardized achievement tests; affective growth on self concept tests; and student coping style, as recorded on the Coping Analysis Schedule for Educational Settings (CASES). Clusters of items from the observation instruments were aggregated as behavioral indications of teacher competency. Reliability data were computed, as well as the correlations between teaching competencies and measures of student growth. A cross-validation technique was also used to combine probabilities from the first and second years of data collection. Results indicated that the observation instruments were reliable indicators of teacher competency; that significant correlations were not consistent from year to year; and that many of the teacher competencies should be renamed. (GDC)