Abstract: Sanitation systems are built to be robust, that is, they are dimensioned to cope with population growth and other variability that occurs throughout their lifetime. It was recently shown that building sanitation systems in phases is more cost effective than one robust design. This phasing can take place by building small autonomous decentralised units that operate closer to the actual demand. Research has shown that variability and uncertainty in urban development does affect the cost effectiveness of this approach. Previous studies do not, however, consider the entire sanitation system from collection to treatment. The aim of this study is to assess the economic performance of three sanitation systems with different scales and systems characteristics under a variety of urban development pathways. Three systems are studied: (I) a centralised conventional activated sludge treatment, (II) a community on site source separation grey water and black water treatment and (III) a hybrid with grey water treatment at neighbourhood scale and black water treatment off site. A modelling approach is taken that combines a simulation of greenfield urban growth, a model of the wastewater collection and treatment infrastructure design properties and a model that translates design parameters into discounted asset lifetime costs. Monte Carlo simulations are used to evaluate the economic performance under uncertain development trends. Results show that the conventional system outperforms both of the other systems when total discounted lifetime costs are assessed, because it benefits from economies of scale. However, when population growth is lower than expected, the source-separated system is more cost effective, because of reduced idle capacity. The hybrid system is not competitive under any circumstance due to the costly double piping and treatment. (C) 2016 Elsevier Ltd. All rights reserved.