Electronic Health Records (EHRs) store multiple patients’ information, including medical history, diagnoses and treatments. Computer-interpretable representation of meanings and intentions in EHRs content might play a major role for decision making, as well as for medical system integration and information recovery. However, there is a lack of suitable representation models to specify the relations between semantic models and illocutions, which reflect the intentions of medical content producers. In this paper, we propose an analysis to understand how illocutions are expressed in EHRs. We aim to identify domain-specific terms to convey the different dimensions in which illocutions are classified. Furthermore, this research develops a model, based on Web ontology description languages, to encode and instantiate the illocutions in the medical domain. Obtained results point out that some illocution types and associated terms are predominant in the analyzed content. They highlight the potentiality of our model to explore illocutions in several computing tasks.