Site selection is a critical factor in the ability to provide safe and healthy environment for refugees. At the outset of an emergency, limited resource and time create large planning uncertainties and current refugee sites’ assessments based on field visit are no longer adequate. The selection of unsuitable locations triggers a number of issues such as environmental degradation, competition for scarce resources and, exploitation of water sources. These can be reduced through a rational refugee allocation decision process. The question that arises is, can the refugee allocation process be supported by a structured decision-making (SDM) approach and more specifically, by a portfolio decision analysis (PDA) model? Focusing on water security criteria, this study presents the initial phases of the SDM-approach intended to ensure a more sustainable allocation of newly-arrived refugees among new and/or existing hosing sites. To this end, a procedural tool for supporting WASH (Water, Sanitation and Hygiene) officers in the site assessment is proposed. This was obtained based on 15 semi-structured interviews and questionnaires carried out in a global stakeholder environment, in order to make the SDM framework applicable to other settings. The generic procedural tool was tested in a case-study in south-western Uganda by applying a portfolio decision analysis model. Local and global data were combined with stakeholder preferences to predict the performance of diverse sets of alternatives. The latter were generated according to different combinations of numbers of refugees, hosting locations and percentage of water extracted from surface water, groundwater and rain water. To identify efficient portfolios, we used the Robust Portfolio Modelling - Decisions software. Results showed that overlooked solutions outperforms over the current allocation strategy. In specific, the scatter of newly arrived refugees showed the highest scores on availability of water, socio-economic costs and host communities' advantages. The proposed framework provides also options for the optimal repartition of the future water extraction among available water sources, aiming to avoid their depletion while preserving sustainable costs of the water services.