Here, we propose a new model for understanding the Warburg effect in tumor metabolism. Our hypothesis is that epithelial cancer cells induce the Warburg effect (aerobic glycolysis) in neighboring stromal fibroblasts. These cancer-associated fibroblasts, then undergo myo-fibroblastic differentiation, and secrete lactate and pyruvate (energy metabolites resulting from aerobic glycolysis). Epithelial cancer cells could then take up these energy-rich metabolites and use them in the mitochondrial TCA cycle, thereby promoting efficient energy production (ATP generation via oxidative phosphorylation), resulting in a higher proliferative capacity. In this alternative model of tumorigenesis, the epithelial cancer cells instruct the normal stroma to transform into a wound-healing stroma, providing the necessary energy-rich micro-environment for facilitating tumor growth and angiogenesis. In essence, the fibroblastic tumor stroma would directly feed the epithelial cancer cells, in a type of host-parasite relationship. We have termed this new idea the "Reverse Warburg Effect." In this scenario, the epithelial tumor cells "corrupt" the normal stroma, turning it into a factory for the production of energy-rich metabolites. This alternative model is still consistent with Warburg's original observation that tumors show a metabolic shift towards aerobic glycolysis. In support of this idea, unbiased proteomic analysis and transcriptional profiling of a new model of cancer-associated fibroblasts (caveolin-1 (Cav-1) deficient stromal cells), shows the upregulation of both (1) myo-fibroblast markers and (2) glycolytic enzymes, under normoxic conditions. We validated the expression of these proteins in the fibroblastic stroma of human breast cancer tissues that lack stromal Cav-1. Importantly, a loss of stromal Cav-1 in human breast cancers is associated with tumor recurrence, metastasis, and poor clinical outcome. Thus, an absence of stromal Cav-1 may be a biomarker for the "Reverse Warburg Effect," explaining its powerful predictive value.