Seven features of valid classroom observations procedures, a five-step measurement schemata, and a systematic classroom observation procedure designed to reduce measurement error are presented. The following features are addressed: (1) the inference of the observation is established a priori and confirmed post hoc; (2) craft knowledge and expertise in pedagogy and teaching is developmental; (3) instrumentation must be built upon a research base, an empirical base, or a combination of the two; (4) observation criteria are generic and content specific; (5) observers are well-trained in the process of observing; (6) observers are well-honed on the meaning of teaching behaviors to be observed; and (7) procedures vary depending upon the inferences that are made. The five steps include observation, recordkeeping, retrieval of observations, analysis, and judgment. A primary issue addressed is the question of the adequacy of traditional measurement procedures for investigating the psychometric properties of classroom observations. Theoretical and empirical evidence illustrates the need for carefully examining traditional alternative methodologies for establishing psychometric properties. Classroom observations may not appear reliable in the true-score sense, but are valid if all seven features are present. Five data tables are included. (TJH)