Emergence of invasive and metastatic behavior in malignant tumors can often lead to fatal outcomes for patients. The collective malignant tumor behavior resulting from the complex tumor-host interactions and the interactions between the tumor cells are currently poorly understood. Progress towards such an understanding necessarily requires an interdisciplinary and collaborative effort. In this paper, we review a state-of-art simulation technique, i.e., a cellular automaton (CA) model which has been developed by the authors over the past few years to investigate microenvironment-enhanced invasive growth of avascular solid tumors. This CA model incorporates a variety of microscopic-scale tumor-host interactions, including the degradation of the extracellular matrix by the malignant cells, nutrient-driven cell migration, pressure build-up due to the deformation of the microenvironment by the growing tumor and its effect on the local tumor-host interface stability. Moreover, the effects of cell-cell adhesion on tumor growth are also explicitly taken into account. A number of bench-mark collective invasion behaviors have been successfully reproduced via the CA model, including the emergence of elongated invasion branches characterized by homotype attraction and least resistance path, development of rough tumor surface in a high-pressure confined environment, as well as reduced invasion due to strong cell-cell adhesion. Such simulated bench-mark behaviors strongly indicate the validity and predictive power of the CA model. In addition, the CA model allows one to investigate the role of various different microenvironment factors in the progression of the neoplasm, in particular, the promotion and enhancement of tumor malignancy. As an example, a “phase diagram” that summarizes the dependency of tumor invasive behavior on extracellular matrix (ECM) rigidity (density) and strength of cell-cell adhesion is constructed based on comprehensive simulations. In this simple phase diagram, a clear transition from non-invasive to invasive behaviors of the tumor can be achieved by increasing ECM rigidity and/or decreasing the strength of cell-cell adhesion. This model, when properly combined with clinical data, in principle enables one to broaden the conclusions drawn from existing medical data, suggest new experiments, test hypotheses, predict behavior in experimentally unobservable situations, be employed for early detection and prognosis, and to suggest optimized treatment strategy for individual patient.