Pressure on freshwater resources is increasingly covered by methodological developments addressing freshwater use in the field of Life Cycle Assessment (LCA). These developments ultimately lead to the publication of the ISO 14046 standard to define the principles, requirement and guidelines for a water footprint in August 2014. The objective of this thesis is to foster the application of water footprint by identifying good practices and developing operational approaches to assess and improve its discriminatory power. Indeed, academic development and LCA application by practitioners often evolve within distinct communities and with various practical constraints. For instance, newly developed methods are spatially differentiated to reduce model uncertainty, improve accuracy, precision and confidence in LCA results. However, models officially recommended by the European commission for product water footprints are generic (i.e., results do not depend on location), such as USEtox for the modelling of the human toxic and ecotoxic impacts. The trade-off between theoretical variability reduction and decrease of uncertainty in practice considering the level of geographical information available to the practitioner requires solutions meeting the needs of both LCA actors. This thesis answers the need to bridge methodological development and water footprint application through three main specific objectives: - Review existing and applicable inventory and impact assessment methods that address quantitative freshwater use in a life cycle perspective and provide preliminary application recommendations for practitioners - Analyse spatial differentiation of toxic emissions to water applied to USEtox at the inter-continental and intra-continental level and explore simplified spatial differentiation approaches such as spatial archetypes - Develop a fate and exposure characterization model and factors for toxic emissions into water with global coverage and fine resolution and test practical solutions to apply spatial characterization factors on a case study The first part of this thesis (Chapter 2) focuses on quantitative aspects of water use by providing a comprehensive review of existing and applicable inventory and impact assessment methods that address freshwater use. This analysis proves that a water footprint is already applicable for experienced practitioners relying on the current state-of-the-art methods. It provides a basis to reference the state-of-the-art of methods addressing water use in LCA already used and cited by several authors in the field of water management and specific water footprint case studies. Practical solutions to operationalize spatial differentiation are tested in Chapter 3 and 4, focusing on impact pathways generated by toxic emissions into water. Spatially differentiated landscapes and models are first developed as a reference to test simplified approaches: (1) landscape parameters are created for USEtox to develop continent and sub-continent specific boxes nested within the world and (2) a fate and exposure characterization model and factors are developed to assess toxic emissions into water with global coverage at 0.5°*0.5° resolution. The results of this work prove that using a generic model is acceptable to cover a low resolution such a continent, country, or sector when the detailed emission location is unknown while a finer resolution is essential for a regional impact score at the watershed or point source level.