The importance of locational determinants in determining rental rates in the office market has long been acknowledged and unquestionably become the significant factor for assessing the rental potential of various property types. Nevertheless, the rapid progressing research that discovers the effect of locational determinants on the rental, limited study that was deeply looked into the provision of location to the rental mainly in Malaysia cases. Thus, this study aims to develop a predictive model for office building rentals based on a comprehensive analysis of locational determinants. To achieve the aims of this study, the objectives was outline which are to identify the determinant factors of location and to analyse the significant determinant factors. Through the application of machine learning, this study captured the intricate connections between rentals and different location-related factors. By leveraging advanced algorithms which are decision trees, random forests, support vector machines and gradient boosted trees, the model can effectively handle diverse datasets, encompassing variables such as proximity to central business districts, access to public transport network, neighbourhood and amenities, and traffic condition. Through rigorous data collection and pre-processing, this study constructs a robust dataset comprising historical rental dataset collected in the city area of Kuala Lumpur, Malaysia acquired from Property Services Department (JPPH) were used to train and validate the predictive model via R-squared performance metrics. The results indicate that proximity to Central Business District (CBD) emerges as a significant determinant with the most contribution to the model’s prediction, with offices located in close proximity commanding higher rentals. This study provides valuable insights into the prediction of office building rentals based on locational determinants, offering a practical tool for stakeholders in the real estate industry.