Elevator Parking System (EPS) is a most popular type of automated parking systems due to its relatively low installation and operation costs. This thesis develops design models for two types of EPS based on a recursive optimization/simulation procedure. With integer non-linear models, we seek the design parameters, which minimize the total cost of the EPS while complying with the desired level of performances. With simulation models, we investigate the dynamic nature of the system and obtain relationships between the design parameters and the performance measures with the multiple linear regression. The validity of the model is examined through case problems.