Estimating the required number of emergency service treatment beds must be sensitive to utilization patterns and strategic operational assumptions. This article describes key issues and illustrates techniques for the analysis of arrival and service times. Seasonal arrival patterns, time of day of arrivals, and common statistical distributions for length of stay are discussed. Alternative modeling approaches to estimate future bed needs are described, including visits/year per treatment space, simple queuing modeling, and detailed computer simulation. Sample estimates of treatment rooms needs are provided for typical arrival rates and lengths of stay. A generalized regression model based on the simulation trials is suggested for cases that fall outside of the illustrated simulation case studies.