Aggressive breast cancer has been shown to shift its metabolism towards increased lipid catabolism as the primary carbon source for oxidative phosphorylation. In this study, we present a technique to longitudinally monitor lipid metabolism and oxidative phosphorylation in pre-clinical tumor models to investigate the metabolic changes with mammary tissue development and characterize metabolic differences between primary murine breast cancer and normal mammary tissue. We used optical spectroscopy to measure the signal of two simultaneously injected exogenous fluorescent metabolic reporters: TMRE (oxidative phosphorylation surrogate) and Bodipy FL C16 (lipid catabolism surrogate). We leverage an inverse Monte Carlo algorithm to correct for aberrations resulting from tissue optical properties and to extract vascular endpoints relevant to oxidative metabolism, specifically oxygen saturation (SO2) and hemoglobin concentration ([Hb]). We extensively validated our optical method to demonstrate that our two fluorescent metabolic endpoints can be measured without chemical or optical crosstalk and that dual measurements of both fluorophores in vivo faithfully recapitulate the measurements of each fluorophore independently. We then applied our method to track the metabolism of growing 4T1 and 67NR breast tumors and aging mammary tissue, all highly metabolic tissue types. Our results show the changes in metabolism as a function of mammary age and tumor growth, and these changes can be best distinguished through the combination of endpoints measured with our system. Clustering analysis incorporating both Bodipy FL C16 and TMRE endpoints combined with either SO2 or [Hb] proved to be the most effective in minimizing intra-group variance and maximizing inter-group differences. Our platform can be extended to applications in which long-term metabolic flexibility is important to study, for example in tumor regression, recurrence following dormancy, and responses to cancer treatment.