This research used data from educators and teacher candidates in Ohio (a state in the Midwestern United States) to produce a valid and reliable instrument for measuring instructional coaching skills. The research involved three pilot tests with large samples of respondents. Initial items came from the extant literature on instructional coaching, and data analyses enabled refinement of these initial items. Principal component analysis revealed that the final set of items loads strongly on one factor, with the factor structure remaining consistent across samples. Reliability measures showed that the instrument has strong internal consistency. As a technically adequate measure, the instrument can be used by educators to measure their own instructional coaching skills or as a pre-assessment and post-assessment for professional development programs intending to improve instructional coaching skills.