Summary. This paper reports one of a series of studies conducted to investigate the role of biomedical knowledge in clinical reasoning. It was motivated by findings from our earlier studies that demonstrate that when specific basic science information is provided to medical students prior to solving a clinical case, they are unable to use this information in explaining the patient problem. An experiment was designed to investigate the use of biomedical information in the explanation of a clinical problem without any basic science information (spontaneous explanation) and where basic science information was provided after the clinical case (biomedically primed explanation). The results are discussed in the context of a two‐stage model of diagnostic reasoning. The first stage is referred to as data‐driven reasoning, and is characterized by the triggering of inferences from observations in the data to hypotheses. The second stage is designated as predictive reasoning, and is characterized by the generation of inferences driven by hypotheses. The results show that, with the exception of final‐year medical students, the use of biomedical information interfered with the data‐driven reasoning process. However, it did facilitate the process of predictive reasoning by the students. It is proposed that a sound disease classification scheme is necessary before biomedical knowledge can facilitate both data‐driven and predictive reasoning during clinical problem‐solving.