To develop and validate a surgical performance indicator based on severe urinary complications that require an intervention within 2 years of radical prostatectomy (RP), identified in hospital administrative data. Men who underwent RP between 2008 and 2012 in England were identified using hospital administrative data. A transparent coding framework based on procedure codes was developed to identify severe urinary complications which were grouped into 'stricture', 'incontinence' and 'other'. Their validity as a performance indicator was assessed by evaluating the consistency with diagnosis codes and association with patient and surgical characteristics. Kaplan-Meier methods were used to assess time to first occurrence and multivariable logistic regression was used to estimate adjusted odds ratios (ORs) for patient and surgical characteristics. A total of 17 299 men were included, of whom 2695 (15.6%) experienced at least one severe urinary complication within 2 years. High proportions of men with a complication had relevant diagnosis codes: 86% for strictures and 93% for incontinence. Urinary complications were more common in men from poorer socio-economic backgrounds (OR comparing lowest with highest quintile: 1.45; 95% confidence interval [CI] 1.26-1.67) and in those with prolonged length of hospital stay (OR 1.54, 95% CI 1.40-1.69), and were less common in men who underwent robot-assisted surgery (OR 0.65, 95% CI 0.58-0.74). These results show that severe urinary complications identified in administrative data provide a medium-term performance indicator after RP. They can be used for research assessing outcomes of treatment methods and for service evaluation comparing performance of prostate cancer surgery providers.