Surface-functionalized Fe3O4 nanoparticles are emerging as promising agents for the selective and magnetic separation of various biological molecules. In principle, by engineering the surface of Fe3O4 nanoparticles, they can be applied as tracers to seek and recognize metabolites and secretions of specific diseases. In this report, we developed Fe3O4 nanoclusters with high magnetization and an amino-functionalized surface via the reaction between FeCl2, a hydrazine reductant, and a gelatin polymer to demonstrate magnetically separated prevalent urinary crystals. The surface of the gelatin-coated Fe3O4 nanoclusters was modified by using aminopropylphosphonic acid by amine coupling using EDC and NHS, which led to the exposure of their phosphonic acid groups and improved their affinity for fine Ca-based urinary crystals in the patient's urine. By subjecting the Fe3O4 nanoclusters that were bound to urinary crystals to Raman spectroscopy analysis, the crystalline types of the pre-concentrated urinary components were easily identified. The assignment of the vibration peaks of the crystals is promising for eliminating the false positives that occur when using a microscopic analysis method for urine crystal diagnosis. Sample preparation and identification required less than 10 min. Finally, we demonstrated that this non-invasive analytic platform exhibits a rapid and efficient detection rate of single- and multi-component urinary crystals from urine metabolites. A good correlation (86%) was observed between this non-invasive analytic platform and the diagnostic reports from 35 urolithiasis patients. We expect that this Fe3O4 nanocluster integrated Raman spectrum method will provide crystal information that could help early management for urolithiasis patients.