Abstract: Objective To establish a proteomic pattern for colorectal cancer(CRC) screening by comparing serum proteomic spectra between CRCs and healthy individuals. Methods Serum samples were collected from 80 CRCs and 78 healthy volunteers, and randomly divided into model construction group and validation group. The weak cation exchange(WCX) beads combined matrix-assisted laser desorption/ionization time of flight mass spectrometry(MALDI-TOF MS) technique were used to detect the mass spectrometry data. The obtained mass spectrometry data were then analyzed using Clinprot Tools 2.2 software. A model identifying CRC from healthy volunteers was built in the model construction group and evaluated in the model validation group for reliability. Serum CEA from validation group was detected using ELISA kit as a control. Results Forty-four differentially expressed proteins in serum(P<0.05), including 35 up-regulated proteins and 9 down-regulated proteins, were screened by comparing serum protemic spectra between CRCs and healthy volunteers. Three proteins at m/z 1330.95, 2883.96 and 9294.14 were obtained for developing a Clinprot model which could identify CRCs from healthy volunteers with an accuracy of 94.87%(74/78). In a double blind validation, the Clinprot model yielded a sensitivity of 87.50% and a specificity of 89.47%, which surpassed to CEA. Conclusion Clinprot model constructed using MALDI-TOF-MS combined with WCX kit technology allows identifying CRCs from healthy volunteers with high accuracy, which may contribute to the screening of CRCs.