The aim of this work is to develop a nanoparticle-based methodology to find out biomarkers of diagnostic for knee osteoarthritis, KOA, through the use of matrix assisted laser desorption ionization time-of-flight-based mass spectrometry profiling. Urine samples used for this study were obtained from KOA patients (42 patients), patients with prosthesis (58 patients), and controls (36 individuals) with no history of joint disease. Gold-nano particle MALDI-based urine profiling was optimized and then applied over the 136 individuals. Jaccard index and 10 different classifiers over MALDI MS datasets were used to find out potential biomarkers. Then, the specificity and sensitivity of the method were evaluated. The presence of ten m/z signals as potential biomarkers in the healthy versus non-healthy approach suggests that patients (KOA and prosthesis) are differentiable from the healthy volunteers through profiling. The automatic diagnostic study confirmed these preliminary conclusions. The sensitivity and the specificity for the urine profiling criteria here reported, achieved by the C4.5 classifier, is 97% and 69% respectively. Thus, it is confirmed the utility of the method proposed in this work as an additional fast, non-expensive and robust test for KOA diagnostic. When the proposed method is compared with those used in common practice it is found that sensitivity is the highest, thus with a low false negative rate for diagnostic KOA patients in the population studied. Specificity is lower but in the range accepted for diagnostic objectives.