Food supply chain waste is consisted not only by the visible part generated in commercial and consumer steps, but also by the residues generated during harvesting and processing of food-related plants. A great part of these materials is unavoidable, meaning that they cannot be prevented trough waste reduction policies currently being pushed to achieve better food distribution and end hunger across the globe. Valorising this waste stream may help increasing stability in the food sector, having not only a financial gain, but also an environmental and social positive impact. Mango Processing Waste (MPW) is one of the examples of how such residues are poorly explored, as its large volumes contain several valuable substances, such as bioactive compounds, that can be used in the food industry as ingredients and as nutraceuticals, pesticides in agriculture, biocides and other uses. The flavonol hyperoside and the xanthone mangiferin are the main secondary metabolites found in MPW, being reported to have biological activities that range from antioxidant to pesticidal and pharmacological potential uses. In a broader context considering the use of MPW in a biorefinery and circular economy concepts, sustainable processes are required to meet future sustainability parameters. Therefore, Microwave-Assisted Extraction (MAE) and Matrix Solid-Phase Dispersion (MPSD), two sample preparation techniques, were discussed and studied as proposed green and sustainable methodologies for the extraction of mangiferin and hyperoside from MPW. Doehlert and Box-Behnken experimental designs were used to help assessing the influences of the variables of each technique, allowing to employ a Response Surface Methodology (RSM) to visualize the best conditions and calculate optimum parameters. Fast extraction was achieved using MAE, which obtained maximum response of 261.39 mg kg−1 of mangiferin and 244.44 mg kg−1 of hyperoside. Higher yields were obtained using MSPD methodology, with an extraction yield of 352.90 mg kg−1 and 398.52 mg kg−1 of mangiferin and hyperoside, respectively. The conditions that allowed maximum simultaneous extraction concentrations were calculated using the desirability function. MAE and MSPD methodologies were compared, with an overall conclusion that both were adequate for the determination of the two analytes and can be further studied to be used in higher scales.