This paper presents a method for evaluating smart grid functions that are implemented to operate residential loads. The proposed assessment method is developed based on the energy not-served (ENS) determined at a point-of-supply feeding residential loads. Smart grid functions can operate energy storage appliances (household water heaters, air conditioners, and heating units) to store thermal energy during the daily off-peak-demand hours. The stored thermal energy is discharged during the daily peak-demand hours, thus reducing the power demands of residential loads. The differences in daily energy demands created by smart grid functions can provide an accurate assessment of the effectiveness of smart grid functions. The ENS-based method is tested for 200 residential households fed from 4 distribution transformers, and are operated by smart grid functions. In these tests, smart grid functions are implemented by the peak-demand management, direct load control, and demand response. Test results demonstrate the accuracy and simplicity of the ENS-based method to assess smart grid functions in terms of the ability to reduce the power demands of residential loads during peak-demand hours.