This dissertation addresses the growing pressure on Western healthcare systems caused by macroeconomic factors like aging populations and staff shortages. It focuses on decision-making under uncertainty, particularly in hospital capacity management, which is crucial for maintaining care continuity. Conducted in collaboration with Diakonessenhuis, a mid-sized hospital in Utrecht, the research emphasizes practical impact, with several developed models already supporting hospital operations.nnThe thesis is divided into two parts: predictive and prescriptive methods. The predictive part explores forecasting tools to reduce uncertainty. It includes a Bayesian model for detecting sudden changes in medication demand, a hybrid approach combining expert estimates and historical data to forecast bed occupancy, and a Monte Carlo simulation for better planning of operating room schedules. These models have been implemented at Diakonessenhuis, leading to measurable improvements such as saving up to ten beds per day.nnThe prescriptive part focuses on decision-making strategies. It introduces a robust optimization model for scheduling clinical sessions and a simulation-based method for managing surgical instrument inventory. The latter achieved up to 28% inventory savings without compromising care quality.nnThe dissertation concludes that while predictive and prescriptive models hold great promise, successful implementation remains a major challenge. Future research should prioritize practical integration to maximize real-world impact.