Background: Teaching clinical reasoning is a challenge in medical education. Aim: To design a clinical reasoning assessment instrument. Material and methods: Structured interviews were carried out to six physicians with at least five years’ experience. The Grounded Theory method was used to determine the relevant categories of the clinical reasoning process and the modified Delphi expert judgment method to validate the categories, the definition of observable behaviors and the format of the instrument. Results: The relevant reasoning categories were the reason for consultation, medical history, physical examination, additional tests, diagnosis, therapeutic options and reasoning reassessment capacity. Expert judgment assessed at a level of "strongly agree" and "agree" the sufficiency, clarity and pertinence of all categories, related observable behaviors and instrument format. The internal Kappa consistency yielded an index of 0.92. Conclusions: The resulting instrument was constructed with the following axes derived from the main categories and subcategories: reason for consultation, history, physical examination, additional tests, diagnosis, therapeutic options and reassessment capacity.Background: Teaching clinical reasoning is a challenge in medical education. Aim: To design a clinical reasoning assessment instrument. Material and methods: Structured interviews were carried out to six physicians with at least five years’ experience. The Grounded Theory method was used to determine the relevant categories of the clinical reasoning process and the modified Delphi expert judgment method to validate the categories, the definition of observable behaviors and the format of the instrument. Results: The relevant reasoning categories were the reason for consultation, medical history, physical examination, additional tests, diagnosis, therapeutic options and reasoning reassessment capacity. Expert judgment assessed at a level of "strongly agree" and "agree" the sufficiency, clarity and pertinence of all categories, related observable behaviors and instrument format. The internal Kappa consistency yielded an index of 0.92. Conclusions: The resulting instrument was constructed with the following axes derived from the main categories and subcategories: reason for consultation, history, physical examination, additional tests, diagnosis, therapeutic options and reassessment capacity.