Computational modelling methods have been used to predict the risks from lead in drinking water across a simulated supply zone, for a range of plumbosolvency conditions and a range of extents of occurrence of houses having a lead pipe, on the basis of five risk benchmarking methods. For the worst case modelled (very high plumbosolvency and 90% houses with a lead pipe) the percentage of houses at risk in the simulated zone ranged from 34.1 to 73.3%. In contrast, for a simulated phosphate-treated zone and 10% houses with a lead pipe, the percentage of houses at risk in the simulated zone ranged from 0 to 0.4%. Methods are proposed for using computational modelling for different levels of risk assessment, for both water supply zones and individual houses. These risk assessment methods will inform policy, help to set improvement priorities and facilitate a better understanding of corrective options.