The purpose of the current study was to investigate the changes of serum proteome and discover potential biomarkers for Kashin-Beck disease (KBD) using surface-enhanced laser desorption ionization mass spectrometry (SELDI-TOF MS). The serum protein profiles from 102 cases (36 KBD patients, 16 controls in KBD areas, 33 controls in non-KBD areas, and 17 osteoarthritis controls) were detected by SELDI-TOF MS and weak cation-exchange protein chip. Differently expressed peaks in KBD were identified by comparing the data among the four groups using the nonparametric Mann-Whitney test with Bonferroni correction at a significance level of 0.05. Then, those 102 cases were used to generate a classification tree as the training set, and an additional 34 cases were collected as the test set. A classification tree was generated by Biomarker Patterns Software (Ciphergen). Multiple protein changes were detected in the KBD group, including three potential biomarkers (15 886, 5336, 6113 m/z). A classification tree with three distinct proteins was generated. The classification tree was able to distinguish the KBD patients from the controls with 88.89% specificity and 86.36% sensitivity. The study demonstrates that marked serum proteomic changes exist in KBD. The proteins represented by the differently expressed peaks are candidate biomarkers for KBD.