Protein discrimination is increasingly studied because of its roles in biological chemistry. In this study, we introduced a visual sensor using the metal nanoparticles (NPs) of Au, Ag, Cu, Ni and Co to determine the fluorescence (FL) intensity and colour changing patterns to identify ten types of proteins in a polyacrylamide gel. The synthesised NPs have little to no fluorescence, and the addition of proteins immobilised by the gel could differentially enhance the FL intensity and change the FL colour of the NPs, thereby producing FL recognition patterns. Based on the different FL enhancement indices of metal NPs with proteins, linear discriminant analysis (LDA) is applied to differentiate the proteins. By altering the synthesis method of metal NPs, a sensor with different FL patterns for protein discrimination is created.In addition, we have applied this method to detect the protein mixtures after one-dimensional (1-D) polyacrylamide gel electrophoresis (PAGE), and to evaluate the serum samples from patients with liver diseases and thalassemia and from healthy people. This protein discrimination method is rapid, costeffective and visual, and it has a "turn on" effect for metal NPs, which may be extended to other proteins or more complex bioanalytes. † Electronic supplementary information (ESI) available: Preparation for one-dimensional polyacrylamide gel electrophoresis (1-D PAGE) and metal NPs-based FL imaging, basic properties of proteins, all the uorescence data, preparation of ve colloidal metal NPs, the emission uorescence spectra and the UV-vis absorption spectrum, the differentiation of NPs, quantitative analysis of the proteins, FL change proles of NPs in the presence of gel components, and the discrimination of serum samples by the metal NPs based on gel. See