Chronic kidney disease (CKD) is highly prevalent in domestic cats (Felis catus), and the specie is a potential model for human CKD. Mass spectrometry-based urine proteomic analysis is a promising tool for CKD diagnosis and management in human beings; yet, minimal information is available in cats. This study aims to explore alterations in the urinary proteome in feline CKD and assess its diagnostic value. Urine from 38 cats (14 with surgically induced CKD, 24 healthy) was used as training data. High-performance liquid chromatography- tandem mass spectrometry (HPLC-MS/MS) was used for urine proteomic analysis, and proteins’ quantities were calibrated to a spike-in protein serving as internal standard. Principal component analysis (PCA), logistic regression analysis, and functional enrichment analysis revealed and characterized changes in the urinary proteome in cats with CKD, compared to healthy cats. A random forest model was trained and applied to a second set of cats with naturally occurring CKD and a corresponding control group (total n= 40; 23 cats with CKD, 17 healthy controls). Under receiver-operating-characteristic (ROC) curve analysis, the model gave an area under curve (AUC) of 0.985 (95% CI, 0.959-1.000), outperforming serum symmetric dimethylarginine (SDMA), urine protein-to-creatinine ratio (UPC), and some commonly studied urinary biomarkers (i.e., transferrin, albumin, cystatin c). Retinol binding protein 4 (RBP4), the top contributor of the random forest model, achieved a comparable AUC of 0.977 (95% 0.941-1.000). These data provide proof of concept for mass spectrometry-based urinary proteomic analysis application in feline CKD.