This paper aims to investigate the process performances in Emergency Departments with an innovative data-driven approach. This approach permits to gain insights into the Emergency Departments processes showing the entire patient-flow, deploying the performances in term of time and resources on the activities and flows, and identifying process deviations and critical bottlenecks. Moreover, the use of this methodology in real-time might dynamically provide a picture of the real situation inside the Emergency Department in term of waiting times, crowding, resources etc., enabling the management of patient demand and resources in real-time. The proposed methodology exploits the process-mining techniques. Starting from the event data inside the hospital information systems, it permits automatically to extract the patient-flows, to evaluate the process performances, to detect process exceptions, and to identify the deviations between the expected and the actual results. The proposed approach was tested with a real case study inside an Emergency Department. This study responds to the need of novel approaches for monitoring and evaluating processes performances in the Emergency Departments, a particular area of attention for the public opinion and national authorities. The novelty of this data-driven approach is the opportunity to timely connect performances, patient-flows and activities. The suggested method supports the hospital managers in performance and process analysis inside the Emergency Departments. Starting from the results provided by this system, it is possible to explore the root causes of deviations, devise improvement actions, and evaluate the effects of corrective actions.