Objectives To identify predictors of renal function preservation, and to compare the global and split renal function outcomes of robot-assisted partial nephrectomy and laparoscopic partial nephrectomy. Methods Demographic, operative and pathological data, as well as renal function outcomes, of 251 patients who underwent laparoscopic (n = 104) and robot-assisted (n = 147) partial nephrectomy between 2008 and 2018 were retrospectively analyzed. Propensity score matching (1:1) was carried out to adjust for potential baseline confounders. Functional outcomes were assessed based on the estimated glomerular filtration rate and dynamic renal scintigraphy (using Tc-99m-mercaptoacetyltriglycine), including renal volumetric analysis. Results A total of 98 patients were allocated to each partial nephrectomy group. Ischemic (laparoscopic vs robot-assisted partial nephrectomy: 29 vs 15 min, P < 0.001) and operative times (181 vs 100 min, P < 0.001) were shorter in robot-assisted partial nephrectomy. The preservation ratio of global renal function at 3 months (88.3% vs 91.4%, P = 0.040) and 12 months (87.8% vs 91.5%, P = 0.010) postoperatively, and the renal function of the operated kidney (80.3% vs 88.2%, P < 0.001) were greater after robot-assisted partial nephrectomy. In robot-assisted partial nephrectomy, the volume of resected parenchyma was significantly smaller (27.2 vs 15.5 mL, P < 0.001), resulting in greater postoperative normal parenchymal volumes (120 vs 132 mL, P < 0.001) and a greater parenchymal preservation ratio (81.1% vs 90.1%, P < 0.001). The parenchymal preservation ratio was the strongest predictor of renal function preservation after surgery (P < 0.001, odds ratio 6.02). Conclusions Robot-assisted partial nephrectomy allows better preservation of split renal function than laparoscopic partial nephrectomy by increasing the parenchymal preservation ratio. This translates into better postoperative global renal function.