Clinical information system (CIS) whose core component is the electronic medical record (EMR), is critical to efficient health care delivery. However, patient mobility and CIS heterogeneity have strengthened the interoperability problem of EMRs. Although ontology is a panacea to this problem, creating interoperable EMR ontology is still esoteric. This paper explicates how an interoperable EMR ontology was crafted for a tertiary health facility in Nigeria. The ontology named EMRONT was created using the Noy and McGuiness’ methodology and Protégé. EMRONT has been evaluated for quality using competency questions and OntoQA and subsequently refined for interoperability by manual mapping to both Health Level 7 (HL7) standard and SNOMED-CT. EMRONT has been further refined for interoperability by semi-automatic mapping to SNOMED-CT using Snoggle. This paper has made exoteric the hitherto esoteric craft of building interoperable EMR ontology and exposed how semi-automatic mapping of EMR ontology to SNOMED is done in practice.