Objectives: A common complication after renal transplantation is allograft rejection, which often leads to chronic rejection and eventual graft loss. While renal allograft biopsy continues to be considered the gold standard in the diagnosis of chronic rejection. The development of non-invasive methods for the accurate detection of chronic rejection of renal grafts has thus become of important clinical importance. Methods: NMR-based serum metabolomics was employed for the analysis of serum metabolites in 18 renal allograft recipients with chronic rejection (CABMR) and 28 with Acute rejection (NCABMR). Samples were analysed by 800 MHZ NMR spectrometer. The metabolic profiles and differential metabolites of sera were analysed by multivariate statistical analysis (MSA), including orthogonal partial least squares discriminant analysis (OPLS-DA) methods. Results: The orthogonal projection to latent structures discriminant analysis (OPLS-DA) model resulted in an R2(Cum) of 0.9 and a Q2 (Cum) of 0.54 for Chronic rejection and Acute rejection subjects, respectively. Among the differential unregulated metabolites identified in Chronic rejection, NDMA, Citrate, Pyridoxin, NDMA, 3-HB, and proline were upregulated from MSA. NDMA had the highest discriminatory potential (AUC 0.84, P=0.0006) followed by Citrate (AUC 0.79, P=0.02), Proline (AUC 0.74, P=0.01), 3-HB (AUC 0.73, P=0.02) and Pyridoxin (AUC 0.72, P=0.05). The results demonstrated that Chronic rejection possesses an active Phenylalanine, Tyrosine, and Tryptophan biosynthesis pathway. Conclusions: Despite being in its early stages, metabolomics monitoring in kidney transplantation can provide reliable indicators of chronic kidney injuries and allograft rejection The diagnostic model that evolved in this study may prove valuable as a tool for a definitive diagnosis of Chronic rejection and Acute rejection patients after validation in larger sample sizes.