Peaks in patients’ demand for inward hospitalization usually lead to disruptions in the provision of healthcare, having negative effects on patient and staff satisfaction. The two main sources of ward bed demand are the emergency department and the surgical center; while the former is random by nature, the latter may be managed through proper allocation of surgical specialties to time slots (or blocks) in the center’s timetable (or Master Surgical Schedule – MSS), and efficient scheduling of surgical procedures within time slots across specialties. We propose a three-step method to design an MSS timetable. In step 1, we mine historical data to determine the average duration of surgical procedures and the average length of stay in wards required by each surgical specialty. In step 2, we use a genetic algorithm to determine a good quality timetable that minimizes the ward bed demand variability overall specialties. In step 3, we approximate the new timetable to the one currently in use at the hospital through a refinement heuristic. Our propositions were tested using data from a tertiary public teaching hospital. The resulting timetable reduced post-operative ward bed demand variability by 99.9%, keeping 97% of surgical specialties allocated in their original slots. To the best of our knowledge, this is the first method for long-term MSS design that reduces post-operative ward bed demand variability and changes in allocations in the current surgical center’s timetable. We innovate by considering the hospital's current timetable to search for solutions promoting minimum changes to the surgical center’s operation.