This project was completed for a local heating and air conditioning company in Wisconsin, Company X. Company X hypothesized that weather impacted its sales and wanted to predict company sales based on different weather conditions. Four predictive models were chosen to predict the company sales of filters, air conditioners, and furnaces and the time spent on no-heat and no-air conditioning service calls. The models chosen were k-nearest neighbors, artificial neural networks, random forests, and multiple linear regression. Some of the weather variables used in making predictions included daily maximum, minimum, and average temperatures, precipitation amount, snowfall amount, average wind speed, and fastest 5-second wind speed. Double cross-validation was completed and the best model for each company sale item/service was chosen and used to make predictions. The predictions were evaluated, and important weather variables were examined.