This study investigated the predictive validity of the Teacher Performance Assessment Instruments (TPAI) using pupil gains on teacher-made tests (TMT) as a criterion. The TPAI and the TMT's were administered to 56 elementary and secondary classrooms in Georgia. Results supported the validity of the TPAI with a large number of correlations ranging from .30 to .45. Regression analysis suggested that 25 per cent of pupil learning gain variance was correlative with the successful implementation of classroom procedures and practices. Implications of the methodology for teacher effectiveness research are discussed. (Author)