Determining the optimal number of beds in a hospital unit is often a very critical task. Patients' number of arrivals and length of stay are random variables which necessitates the treatment with the number of patients at hospital as a stochastic process, and thus adding to the complexity of the bed sizing problem. The optimal number of beds is a®ected by some critical parameters such as target utilisation level, admission rate and target service level. This study applies a discrete event simulation model to approximate the true relationships between di®erent control parameters and optimal number of beds. A goal programming model is then solved to ¯nd the optimal number of beds that maintains minimal deviation from target admission and utilisation levels. Data have been collected for the calendar year 2019 and then analysed and used in the simulation model. Mathematical relationships are then embedded in a multiobjective optimisation model that ¯nds the optimal number of beds in an Intensive Care Unit that minimises the deviations from a pre-speci¯ed service and beds utilisation levels.