This paper discusses modeling the price elasticities of substituting peak-to-off-peak energy usage based on demand response programs. Two commonly used price response models are adapted to our data analysis: the Constant Elasticity of Substitution (CES) model and the Generalized Leontief (GL) model. The models were estimated using data on real electricity consumption by an electric utility study group under a dynamic-rate program. The sensitivity of substitution elasticity to weather was also taken into account, as well as customer characteristics such as the presence of electric or gas heating, central air-conditioning, programmable communicating thermostats, and relative income levels. These substitution elasticities can be used to predict the amount of load shifting expected from new rate structures and thus inform the design of a dynamic pricing program.