The purpose of this exploratory, descriptive study was to investigate teachers' perceptions of three types of feedback on students' performance to guide instructional improvements. These include: (1) formative assessment error analyses, (2) mastery charts of class progress on formative assessments, and (3) summative assessment results comparisons with previously taught classes. Self-report survey data from 92, K-12 teachers involved in a pilot mastery learning program revealed that analyses of students' errors on formative assessments were consistently rated the most useful in planning corrective instruction and in making instructional improvements. Mastery charts and summative assessment results were considered more useful in evaluating the overall effectiveness of mastery learning and in revising implementation procedures. Implications for professional learning and program implementation are discussed.