Predicting resource consumption in the built environment and its associated environmental consequences is one of the core challenges facing policy-makers and planners seeking to increase the sustainability of urban areas. The study of land-use change has many implications for infrastructure design, resource allocation, and urban metabolism simulation. While most urban models focus on horizontal growth patterns, few investigate the impacts of vertical characteristics of urbanscapes in predicting land-use changes. In this paper, Building-form variables are introduced as a new determinant factor for investigating effects of vertical characteristics of an urbanscape in predicting land-use change. This work outlines an automated method for generating building-form variables from Light Detection and Ranging (LIDAR) data by using Density-Based Spatial Clustering and normal equations. This paper presents a Land-Use Model that uses Remote Sensing, GIS, and Artificial Neural Networks (ANNs) to predict urban growth patterns within the IUMAT framework (Integrated Urban Metabolism Analysis Tool), which is an analytical platform for quantifying the overall sustainability in the urbanscape. The town of Amherst in Western Massachusetts (for the period of 1971–2005) is used as a case study for testing the model. By isolating the weights of each explanatory variable in models, this study highlights the influence of building geometry on future development scenarios.