Grading of renal cell carcinoma (RCC) has prognostic significance, and there is recent consensus by the International Society of Urological Pathology (ISUP) that for clear cell and papillary RCC, grading should primarily be based on nucleolar prominence. Microscopic tumor necrosis also predicts outcome independent of tumor grading. This study was undertaken to assess whether the incorporation of microscopic tumor necrosis into the ISUP grading system provides survival information superior to ISUP grading alone. Data on 3017 patients treated surgically for clear cell RCC, 556 for papillary RCC, and 180 for chromophobe RCC were retrieved from the Mayo Clinic Registry. Median follow-up periods were 8.9, 9.7, and 8.5 years, respectively. Four proposed grades were defined: grade 1: ISUP grade 1+ISUP grade 2 without necrosis; grade 2: ISUP grade 2 with necrosis+ISUP grade 3 without necrosis; grade 3: ISUP grade 3 with necrosis+ISUP grade 4 without necrosis; grade 4: ISUP grade 4 with necrosis or sarcomatoid/rhabdoid tumors. There was a significant difference in survival between each of the grades for clear cell RCC, and the concordance index was superior to that of ISUP grading. The proposed grading system also outperformed the ISUP grading system when cases were stratified according to the TNM stage. Similar results were not obtained for papillary RCC or chromophobe RCC. We conclude that grading for clear cell RCC should be based on nucleolar prominence and necrosis, that ISUP grading should be used for papillary RCC, and that chromophobe RCC should not be graded.