Despite optimal immunosuppressive therapy, more than 50% of kidney transplants fail because of chronic allograft dysfunction. A noninvasive means to diagnose chronic allograft dysfunction may allow earlier interventions that could improve graft half-life. In this proof-of-concept study, we used mass spectrometry to analyze differences in the urinary polypeptide patterns of 32 patients with chronic allograft dysfunction (14 with pure interstitial fibrosis and tubular atrophy and 18 with chronic active antibody-mediated rejection) and 18 control subjects (eight stable recipients and 10 healthy control subjects). Unsupervised hierarchical clustering showed good segregation of samples in groups corresponding mainly to the four biomedical conditions. Moreover, the composition of the proteome of the pure interstitial fibrosis and tubular atrophy group differed from that of the chronic active antibody-mediated rejection group, and an independent validation set confirmed these results. The 14 protein ions that best discriminated between these two groups correctly identified 100% of the patients with pure interstitial fibrosis and tubular atrophy and 100% of the patients with chronic active antibody-mediated rejection. In summary, this study establishes a pattern for two histologic lesions associated with distinct graft outcomes and constitutes a first step to designing a specific, noninvasive diagnostic tool for chronic allograft dysfunction.