The association of a genetic analysis that could improve the diagnostic accuracy of renal cell tumors in biopsy samples would allow better-informed therapeutic decisions. We performed comparative genomic hybridization (CGH) on an ex vivo fine-needle aspiration (FNA) biopsy and a tumor fragment obtained from 75 patients consecutively diagnosed with renal tumors and subjected to radical nephrectomy. The pattern of genomic changes by CGH was used blindly to classify the renal tumors and the genetic findings were subsequently compared with the histopathologic diagnosis. In particular cases, including in two carcinomas with morphologically distinct tumor areas, we performed FISH with several locus-specific probes, and looked for VHL point mutations, exonic rearrangements, or promoter methylation. CGH was successful in 82.7% FNA biopsies and in 96% tumor fragments, with the former allowing genetic diagnosis in 75% of renal cell tumors. The genetic and the initial histological classification differed in two renal neoplasias, but the genetic diagnosis was confirmed after review. The genetic pattern correctly diagnosed 93.5% of clear cell renal cell carcinomas (RCC), 61.5% of chromophobe RCC, 100% of papillary RCC, and 14.3% of oncocytomas, with the negative predictive value being 93.9, 90.7, 100, and 90.2%, respectively. The positive predictive value and specificity of copy number profiles was 100%. We demonstrate that genetic diagnosis by CGH on FNA biopsies can improve differential diagnosis in patients with kidney tumors.