This study aims to predict the long-term residential electricity demand under various climate change scenarios, including annual temperature increases or decreases of 1°C and 0.5°C, as well as CMIP6 climate models. Rising average temperatures and intensified heatwaves lead to higher cooling demand and increased summer peak loads. To analyze this trend, a data-driven framework based on an annual multivariate regression model is developed, incorporating cooling and heating degree days, the number of consumers, and temporal trends as input variables. Daily temperature data are also used to disaggregate annual consumption into daily and hourly scales, enabling detailed assessment of peak load variations. Model evaluation indicates high accuracy, with coefficients of determination of 0.995 for training data and 0.979 for testing data. The root mean square error is approximately 547.82 kWh less than one percent of annual consumption. The climate scenarios reveal that the cooling degree days (CDD) index is the main driver of increased electricity consumption in Tehran, rising from about 780 units in the base year to 3,000 in the mild warming scenario and over 7,200 in the severe one an increase of 285% to 820%. In contrast, the heating degree days (HDD) index decreases under warmer conditions and has minimal impact on demand. Projections suggest that residential electricity consumption will grow by 75% to 132% by 2050. This framework provides an effective tool for network planning and load management under global warming conditions.