BACKGROUND: The existing literature suggests that the surgical mortality (SM) observed with nephrectomy for localised disease varies from 0.6% to 3.6%. OBJECTIVE: To examine age- and stage-specific 30-d mortality (TDM) rates after partial or radical nephrectomy. DESIGN, SETTING, AND PARTICIPANTS: We relied on 24535 assessable patients from the National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) database. MEASUREMENTS: In 12283 patients, logistic regression models were used to develop a tool for pretreatment prediction of the probability of TDM according to individual patient and tumour characteristics. External validation was performed on 12252 patients. RESULTS AND LIMITATIONS: In the entire cohort of 24535 patients, 219 deaths occurred during the initial 30 d after nephrectomy (0.9% TDM rate). TDM increased with age (≤49 yr: 0.5% vs 50-59 yr: 0.7% vs 60-69 yr: 0.9% vs 70-79 yr: 1.2% vs ≥80 yr: 2.0%; χ(2) trend p<0.001) and stage (0.3% for T1-2N0M0 vs 1.3% for T3-4N0-2M0 vs 4.2% for T1-4N0-2M1; χ2 trend p=<0.001). TDM decreased in more recent years (1988-1993: 1.3% vs 1994-1998: 0.9% vs 1999-2002: 0.7% vs 2003-2004: 0.6%; χ2 trend p<0.001) and was lower after partial versus radical nephrectomy (RN) (0.4% vs 0.9%; p=0.008). Only age (p<0.001) and stage (p<0.001) achieved independent predictor status. The look-up table that relied on the regression coefficients of age and stage reached 79.4% accuracy in the external validation cohort. CONCLUSIONS: Age and stage are the foremost determinants of TDM after nephrectomy. Our model provides individual probabilities of TDM after nephrectomy, and its use should be highly encouraged during informed consent prior to planned nephrectomy.