Clinical reasoning is a fundamental skill required of all physicians. Direct observation is onemethod medical schools use to assess clinical reasoning, where faculty observers rate students based on the student’s interaction with a patient. Variability in how individual faculty members define clinical reasoning, however, can reduce assessment reliability. Understanding how faculty make assessment decisions of student clinical reasoning can improve the reliability and validity of medical school’s assessments. Fourteen UC Davis School of Medicine faculty members completed think-aloud interviews while watching a medical student encounter with a standardized patient. Faculty members were asked to assess the student’s clinical reasoning ability and were not provided any information about the student or the case other than a door note. The faculty were then asked to provide written summative feedback to the student. The think-aloud interviews were video-recorded, transcribed, and analyzed using thematic analysis. The analysis provided five themes about how faculty members assess medical students: student factors, situational factors, assessor factors, integration, and judgment. Additional findings about the ways in which faculty provide students narrative feedback were also noted. The themes together create a model of faculty reasoning, the process by which faculty make assessment decisions about a medical student’s clinical reasoning ability. Faculty assessment decisions are influenced by a number of different factors. The ways in which they process information about the student and the encounter, and then integrate it with their own existing knowledge and experience, is unique to the individual. Understanding this process allows for opportunities to influence factors to improve consistency, and therefore validity.