We aimed to identify protein biomarkers that could rapidly and accurately diagnose osteoporosis patients (OPs) using a highly sensitive proteomic immunoassay. Four-dimensional (4D) label-free proteomics analysis was performed to determine the differentially expressed proteins in serum collected from 10 postmenopausal osteoporosis patients and 6 non-osteoporosis patients. The ELISA method was used to select the predicted proteins for verification. Serum was taken from 36 postmenopausal osteoporosis patients and 36 healthy individuals from normal postmenopausal women. Receiver operating characteristic (ROC) curves were used to determine the diagnostic potential of this method. We validated the expression of these six proteins using ELISA. The CDH1, IGFBP2, and VWF of osteoporosis patients were significantly higher than those of the normal group. PNP was significantly lower than that in the normal group. And using ROC curve calculation, serum CDH1 had a cut-off of 3.78 ng/mL with a sensitivity of 84.4%, and PNP had a cut-off of 944.32 ng/mL with 88.9% sensitivity. These outcomes suggest that serum-level CHD1 and PNP have the potential power as effective indicators for the diagnosis of PMOP. Our results suggest that CHD1 and PNP might be associated with the pathogenesis of OP and would be helpful in diagnosing OP. Therefore, CHD1 and PNP may act as potential key markers in OP.