Simultaneous discrimination of multiple enzymes in clinical diagnosis is significance and remains a great challenge. Herein, we reported a facile synthesis of three ternary transition metal sulfide (TTMS) nanoparticles, i.e. FeCo2S4, CuCo2S4 and FeCu2S4, by using L-cysteine as sulphur source for oxidase discrimination. The CL emission from luminol oxidation by hydrogen peroxide (H2O2) was greatly enhanced by TTMS, with the maximal three orders of magnitude CL signal amplification for FeCo2S4. Hence, an ultra-sensitive CL assay was developed for H2O2 detection down to 0.5 pM. Because H2O2 is generated via biocatalytic reaction between substrate and corresponding oxidase, thus, CL sensor arrays can be developed to discriminate different oxidases. The sensing principle is based on the fact that different substrates/oxidases produced different amounts of H2O2. For each substrate/oxidase couple, the sensor array produces a peculiar CL intensity pattern able to be distinguished via linear discriminant analysis. As a proof-of-concept study, four oxidases, i.e. glucose-, sarcosine-, choline- and cholesterol-oxidase, were selected to demonstrate the array’s discrimination performance and were identified successfully with an accuracy of 100%, and achieved the accurate discrimination of oxidases in serum samples. This work demonstrated the potential of the TTMS-based CL sensor array to discriminate multiple oxidases.