This study was undertaken to determine the incidence, risk factors, and in-hospital outcome of nephropathy requiring dialysis (NRD) after percutaneous coronary intervention (PCI), and to evaluate the role of a weight- and creatinine-adjusted maximum radiographic contrast dose (MRCD) on NRD. Data were obtained from a registry of 16,592 PCIs. The data were divided into development and test sets. Univariate predictors were identified and a multivariate logistic regression model was developed. The MRCD was calculated for each patient as: MRCD = 5 ml x body weight (kilograms)/serum creatinine (milligrams per deciliter). Predictive accuracy was assessed by receiver-operating characteristic curve analysis. In the development set, 41 patients (0.44%) developed NRD with a subsequent in-hospital mortality rate of 39.0%. NRD increased with worsening baseline renal dysfunction. Other risk factors included peripheral vascular disease, diabetes mellitus, congestive heart failure, and cardiogenic shock. There was a direct relation between the number of risk factors and NRD. After adjustment for baseline risk factors, MRCD was the strongest independent predictor of NRD (adjusted odds ratio 6.2, 95% confidence interval 3.0 to 12.8). NRD and in-hospital mortality were both significantly higher in patients who exceeded the MRCD compared with patients who did not (p <0.001). In conclusion, NRD following PCI is a rare complication with a poor prognosis. Baseline clinical characteristics identify patients at greatest risk for NRD. Optimization of procedural variables such as timing of the intervention relative to the diagnostic catheterization, staging coronary procedures, or dosing within the MRCD may help reduce the risk of this complication in high-risk patients. A risk prediction tool for NRD with guidelines for prevention is presented.