Considering the current growing interest in new and improved enzymes for use in a variety of applications, the present study aimed to characterize a novel detergent-stable serine alkaline protease from the extremophilic actinobacterium Microbacterium metallidurans TL13 (MmSP) using a combined in silico and experimental approach. The MmSP showed a close phylogenetic relationship with high molecular weight S8 peptidases of Microbacterium species. Moreover, its physical and chemical parameters computed using Expasy's ProtParam tool revealed that MmSP is hydrophilic, halophilic and thermo-alkali stable. 3D structure modelling and functional prediction of TL13 serine protease resulted in the detection of five characteristic domains: [catalytic subtilase domain, fibronectin (Fn) type-III domain, peptidase inhibitor I9, protease-associated (PA) domain and bacterial Ig-like domain (group 3)], as well as the three amino acid residues [aspartate (D182), histidine (H272) and serine (S604)] in the catalytic subtilase domain. The extremophilic strain TL13 was tested for protease production using agricultural wastes/by-products as carbon substrates. Maximum enzyme activity (390U/gds) was obtained at 8th day fermentation on potato peel medium. Extracellular extract was concentrated and partially purified using ammonium sulfate precipitation methodology (1.58 folds purification fold). The optimal pH, temperature and salinity of MmSP were 9, 60°C and 1M NaCl, respectively. The MmSP protease showed broad pH stability, thermal stability, salt tolerance and detergent compatibility. In order to achieve the maximum stain removal efficacy by the TL 13 serine protease, the operation conditions were optimized using a Box-Behnken Design (BBD) with four variables, namely, time (15-75min), temperature (30-60°C), MmSP enzyme concentration (5-10U/mL) and pH (7-11). The maximum stain removal yield (95 ± 4%) obtained under the optimal enzymatic operation conditions (treatment with 7.5U/mL of MmSP during 30min at 32°C and pH9) was in good agreement with the value predicted by the regression model (98 ± %), which prove the validity of the fitted model. In conclusion, MmSP appears to be a good candidate for industrial applications, particularly in laundry detergent formulations, due to its high hydrophilicity, alkali-halo-stability, detergent compatibility and stain removal efficiency.