Instructional time use is an intervention without equal. The measure of such has clear and important implications for special education practice and research. Although exhortations to maximize instruction and thereby student engagement exist throughout the literature, few studies discuss how special education teachers use their time, and none address the sampling or measurement issues related to differences of time across the academic calendar, by day or by teacher. This empirical investigation reports the requirements for adequate sampling, discussing variance by teacher and across the calendar year with attention to the standard error of measurement and denoting when stability is achieved. High-quality measurement provides opportunity for a more scientific approach to maximizing time use and thus student achievement. Results suggest that approximately 10 days appear to be sufficient to obtain a reasonably stable measure of teacher time use with a multiple code instrument designed to capture a variety of teacher behaviors. (Contains 4 tables and 4 figures.)