Objective: To construct a preoperative evaluation system for partial nephrectomy using CT three-dimensional visualization technology and to explore its practical value. Methods: The clinical data of the patients who underwent partial nephrectomy for renal tumors in Department of Urology, Peking University First Hospital were collected retrospectively. At the same time, the homogenized standard data of patients who underwent partial nephrectomy for renal tumors were collected in 16 clinical centers in China. The CT three-dimensional visualization system was applied (IPS system, Yorktal) to evaluate tumor anatomy, blood supply, perirenal fat and other information. The parameters were summarized to build a three-dimensional nephrometry system, on the basis of which virtual surgery design and intraoperative navigation were completed. Results: A three-dimensional visualization image was established based on the enhanced CT urography. The nephrometry system included the longest diameter and volume of the tumor, proportion volume of tumor invading the parenchyma, maximum depth of the tumor invading the parenchyma, contact surface area, flatness of the tumor surface, renal segment where the tumor was located, vascular variation, and perirenal fat. The average two-dimensional diameter of the tumor was (2.78±1.43) cm, the average three-dimensional maximum diameter was (3.09±1.35) cm, and the average postoperative pathological size was (3.01±1.38) cm. The maximum tumor diameter in the three-dimensional image was significantly related to the prolonged renal artery clamping time and intra-operative blood loss (r=0.502, P=0.020; r=0.403, P=0.046). The three-dimensional and pathological tumor volume were (25.7±48.4) cm³ and (33.0±36.4) cm³, respectively (P=0.229). The tumor volume was significantly related to the intraoperative blood loss (r=0.660, P<0.001). The proportion volume of the tumor invading into renal parenchyma was significantly related to the prolongation of renal artery clamping and the occurrence of postoperative complications (r=0.410, P=0.041; r=0.587, P=0.005). The tumor contact surface area and the presence of vascular variation did not show correlation with the perioperative data and postoperative complications. While the preoperative evaluation was completed, the reconstructed three-dimensional image could be zoomed, rotated, combined display, color adjustment, transparency, and simulated cutting on the Touch Viewer system. The process generally consisted of showing or hiding the tissue, adjusting the transparency of the interested area, rotating and zooming the image to match the position of the surgical patient. Together, these functions met the requirements of preoperative virtual surgery plan and intraoperative auxiliary navigation. Conclusion: Three-dimensional images can provide a more intuitive anatomical structure. The CT three-dimensional visua-lization system clearly displays tumor anatomical parameters, blood supply and perirenal fat. The three-dimensional nephrometry system for renal tumors can help predict the difficulty of partial nephrectomy and perioperative complications. Importing the reconstructed three-dimensional visualization image into the specified program or robot operating system can complete virtual surgery and intraoperative navigation, helping the surgeon to better grasp the surgical process. The indexes included in the nephrometry system and the score weights of each index need to be confirmed and perfected by multi-center study with large samples.