Landslide susceptibility modeling is an essential early step towards managing landslide risk. A minimum of $4.8 million is lost due to landslide related damages every year in Illawara region of Australia. At present, Data mining and knowledge discovery techniques are becoming popular in building landslide susceptibility models due to their enhanced predictive performances. Until now, the lack of tools to undertake data extraction and making the predictions have limited the applicability of this novel technique in landslide model building. This paper discusses the development of the LSDM (Landslide Susceptibility Data Mining) toolbar which was designed to utilize machine learning techniques within a GIS environment by coupling GIS and data mining software (See5) capabilities. The software development kit available with ArcGIS v.10 has been utilized in developing the toolbar add-in. The fundamental tasks; data preparation, model optimizing, derivation of decision trees, predictions and validation are all performed using the individual controls available in the toolbar. This tool automates the entire model building process and in preparation of training data and producing outcomes that are compliant with both national and international Landslide Risk management guidelines.