Current methodologies for Information Systems (ISs) audits suffer from some limitations that could question the effectiveness of such procedures in detecting deviations, frauds, or abuses. Process Mining (PM), a set of business-process-related diagnostic and improvement techniques, can tackle these weaknesses, but literature lacks contributions that address this possibility concretely. Thus, by framing PM as an Expert System (ES) engine, this paper presents a five-step PM-based methodology for IS audits and validates it through a case in a freight export port process managed by a Port Community System (PCS), an open electronic platform enabling information exchange among port stakeholders.The validation pointed out some advantages (e.g. depth of analysis, easier automation, less invasiveness) of our PM-enabled methodology over extant ESs and tools for IS audit. The substantive test and the check on the PCS processing controls and output controls allowed to identify four major non-conformances likely implying both legal and operational risks, and two unforeseen process deviations that were not known by the port authority, but that could improve the flexibility of the process. These outcomes set the stage for an export process reengineering, and for revising the boundaries in the process flow of the PCS.