Abstract: This project analyses weather and pest-prevalence data in South Africa to identify possible relationships between changing climatic patterns and pest populations using machine learning techniques. Three predominant datasets were received from the Forestry and Agricultural Biotechnology Institute (FABI) for contributing to this project, namely: Temperature and rainfall records of roughly 6000 weather stations across South Africa, Sirex noctilio (Sirex) pest inspections, and Leptocybe invasa (Leptocybe) pest inspections. The weather and pest datasets provided had no linking attributes, other than spatial references. The data was processed using a feature engineering algorithm to link the weather and pest datasets and prepare the data for modelling. The engineered parameters were then used to train XGboost and Support Vector Machine (SVM) models. Thereafter, the models were tested for stability and prepared for deployment. Deployment allows the user to input weather parameters to determine a pest presence prediction probability at the user-defined location and across South Africa. The XGboost model achieves 81% accuracy for Leptocybe and 64% accuracy for Sirex. Overall, XGboost performs 10 to 15% better than the SVM when classifying both pests. However, both models classify Leptocybe more accurately than Sirex.