As the field of higher education has used several methods to assess college teaching, the validation of these measures and how they relate to each other has been sparse. This study drew on data from a multi-institutional study of 587 college courses to examine the efficacy of two methods of measuring college teaching (class observation and syllabus analysis) in understanding six college teaching constructs. Results of the multi-trait multi-method (MTMM) modeling analysis revealed a confluence of theory and method in validation: observation was better suited for situated teaching theories that are embedded within disciplines and learners, while syllabus analysis was stronger for robust constructs (such as time on task) that examine coursework across contexts.