This project studied the impacts of six different factors (land cover, soil type, soil hydraulic, slope, precipitation, and temperature) on the N and P levels in streams that make up the Lake Erie watershed and how these factors can be used to predict the N and P levels of unobserved areas. The selected Random Forest model does an impressive job at accurately predicting expected values. The error metrics show a large improvement over an ordinary least squares regression model. The Random Forest model shows that temperature, precipitation, agricultural land cover, natural land cover, and slope are important factors in predicting values of N and P in the streams around the Lake Erie watershed and that calculated watersheds can be useful in predicting potential areas with high concentrations of these elements. This first pass analysis can be used as a starting point in identifying areas to focus further research or mitigation efforts. It can also be used to inform policies by identifying problem areas and important factors to manage in order to decrease N and P inputs into streams. The methodology used in this study could be used as the basis for future studies using more detailed data about the characteristics of the watersheds used (for example breaking down the landcover classifications into subclasses) and using more observations. With more detailed data and monitoring points, the model and associated prediction values could be potentially improved.