In this paper, statistical modeling of the yearly residential energy demand in Nigeria is presented. Specifically, two statistical models were used, namely, the quadratic regression model without interaction and multiple linear regressions with one period l lagged dependent variable. In the study, 46 years data on the yearly residential energy demand in Nigeria was used along with data on temperature and population which are the explanatory variable. The details of the models development process were presented and the regression coefficients were derived based on the case study dataset. The models prediction performances were assessed in terms of R-square value, the sum of square of error (SSE) and the root mean square error (RMSE). The results showed that the R-square value for the multiple linear regression model with one period lagged dependent variable was 0.8525 (85.25%), SSE was 571850 and RMSE was 111.4970. Also, the R-square value for the quadratic regression model without interaction was 0.7265 (72.65%), SSE was 1060100 and RMSE was 151.8083. The multiple linear regression model with one period lagged dependent variable had a better prediction performance. As such it was used to forecast the yearly residential energy demand in Nigeria for the next eleven years (2018-2028). The forecast result shows that in 2028 the yearly residential energy demand in Nigeria will be 12050.5 MW/h. The results in this study were compared with those obtained in a previous research where only eight years data was used. In that study, the quadratic regression model without interaction was found to be more accurate in the residential energy prediction with very high R-squared value. However, it can be concluded in this study that model prediction performance based on small sample is deceptive; sufficient data is required for effective modeling.