Clinical decision-making (CDM) is key in learning to be a doctor as the defining activity in their clinical work. CDM is often portrayed in the literature as similar to 'trail blazing'; the doctor as the core agent, clearing away obstacles on the path towards diagnosis and treatment. However, in a fieldwork of young doctors in Denmark, it was difficult connect their practice to this image. This paper presents the exploration of this discrepancy in the heart of medical practice and how an alternative image emerged; that of a 'jam session'. The exploration is represented as a case-based hypothesis-testing: first, a theoretically and empirically informed hypothesis (H0) of how doctors perform CDM is developed. In H0, CDM is a stepwise process of reasoning about clinical data, often influenced by outside contextual factors. Then, H0 is tested against a case from ethnographic fieldwork with doctors going through internship. Although the case is chosen for characteristics that make it 'most likely' to verify the hypothesis, verification proves difficult. The case challenges preconceptions in CDM literature about chronology, context, objectivity, cognition, agency, and practice. The young doctor is found not to make decisions, but rather to participate in CDM; an activity akin to the dynamics found in a jam session. Their participation circles in and through four concurrent interrelated constructions that suggest a new conceptualization of CDM; a starting point for a deeper understanding of actual practice in a changing clinical environment.