Ceramic-metal matrix composites produced by powder metallurgy provide a solution in engineering applications where materials with high wear resistance are required. In the mining industry, the wear of materials is a crucial and widely recognized industrial problem as over 50 % of components fail as a result of wear damage. Increasing the wear resistance of these components will contribute to a reduction in maintenance and thereby increase efficiency. In this present research, SS316L-50wt.% Al2O3 composites were fabricated using the powder metallurgy route. The effects of the powder metallurgy processing parameters were studied. The produced cermet composites were characterized with respect to microstructure, density, hardness and toughness. Furthermore, the wear behavior of the composites was studied using pin-on-disc testing under dry sliding conditions. The produced test results were used to improve existing wear models, particularly the Wayne’s model. The highest hardness of 1085.2 HV, the highest density of 94.7 % and the lowest wear rate of 0.00397 mm3/m were obtained at a milling speed of 720 rpm, a compaction pressure of 794.4 MPa and sintering at 1400 °C in an argon atmosphere. Compared to commercial SS316 and fabricated SS316L, the composites had 7.4 times and 11 times lower wear rate, respectively. However, it is shown that using better densification methods such as hot isostatic pressing (HIP) or hot pressing can further substantial enhanced densification and improve of the composites wear resistance. Similar to its effects of the strength and the toughness, the remaining porosity was found to substantially affect the wear resistance of the sintered composites. Therefore, the porosity was used to correct the abrasion parameter in the first step of wear model improvement. The porosity represented a further consideration of the microstructure in addition to the reinforcement particle size introduced earlier by Wayne. In a second model improvement step, the test conditions were introduced in the wear resistance calculation. This model allowed the prediction of corrected wear resistance values that are characteristic of the individual test materials and are widely independent of wear test conditions. The coefficient of correlation of the model was 0.91 with respect to Wayne's data and wear test results from this study, and was 0.66 after generalization to a large range of wear data measured on multiple materials tested under varying test conditions. This opens a potential avenue for a model-based assessment of the wear resistance of novel materials as well as changes that can be expected under different wear conditions.