This paper aims to collect and encapsulate data regarding the modelling efforts made to predict the energy consumption of residential buildings using a prominent statistical method, to help in prediction the subsequent conservation of energy (electricity) consumed in the residential building sector. Among the various statistical methods available linear regression analysis is considered a good option in case of availability of historic building use data via smart metering techniques as it provides reasonably accurate results with a simple approach. In this study, linear and multiple regression analysis were conducted on data collected from a residential building so as to obtain the best model for prediction purposes. The average temperature parameter emerged as the most important predictor variable and relative humidity as the least having low or negative correlations. Also, the study of the cumulative effect of all the independent parameters on the building use variables proved that it is best to incorporate the effects of all the independent variables to obtain accurate results.