A quantitative assessment of rock mass behaviors around underground excavations is essential in mining since it can assist engineers in selecting appropriate mining methods and implementing reliable ground control measures. In this paper, the interactions between the factors affecting the rock mass behaviors around underground excavations are quantified using the Rock Engineering Systems (RES) and Artificial Neural Network (ANN) approaches. To this end, the ground behavior index (GBI) is developed. The RES is applied as a practical tool for determining complex and highly nonlinear correlation among the input parameters via the interaction matrices while ANN is implemented to objectively assign weights to the input parameters of the GBI. Fall of ground (FoG) of rock mass surrounding the excavation comprising gravity induced structurally controlled, block movement and stress induced failure cases were investigated and a comprehensive database on the FoG characteristics was compiled. Several parameters related to the FoG including the rock mass characteristics, the excavation geometry, the excavation supports, the mining methods and the FoG size, were selected to establish the GBI. The Bamangwato Concession Limited (BCL), an underground mine located in Selibe-Phikwe, Botswana was used as case study to compute the proposed GBI. Overall, the validation results showed excellent agreement between the GBI and the field observations. It was concluded that the GBI could be used to provide engineers with reliable quantitative information on the fall of ground as well as the corresponding the hazard level.