Computational modelling has been used in plumbosolvency control optimisation studies involving three English water companies. A zonal model simulates lead emissions across an entire area of supply; it utilises a single pipe model that simulates the lead concentration in drinking water over a 24 h period. A sampling model then simulates the results of random daytime sampling and the results are compared with those obtained by the water company, as a means of validating the zonal model. The simulated zone is created by the random ascription of a range of variables (lead pipe length, pipe diameter, water use, pattern of flow events, etc.), based on the Monte Carlo method. The conditions of the zones that were investigated were wide ranging, and included both low and high alkalinity waters, before and after the dosing of ortho-phosphate to reduce plumbosolvency. Excellent validation was achieved, which enabled the models to be used with confidence for predictive purposes. To achieve the optimisation criteria set by the Drinking Water Inspectorate in 2001, plumbosolvency reductions of between 65 and 91% were found to be required. The minimum length of the 12 mm internal diameter lead pipe that was predicted to continue to fail the 10 μg/L standard, once phosphate dosing had been optimised, ranged from 40 to 50 m, providing a possible basis for selective replacement. Cross-referencing the zonal modelling results to laboratory treatment tests indicated that the optimum average ortho-phosphate dose varied from 0.9 to 2.0 mg/L (P), dependent on water-specific requirements.