The structure and function of estuarine food webs change in response to both natural and anthropogenic stresses. The construction of quantitative food webs and their analysis by means of Ecological Network Analysis provides outputs that have been used in many studies to assess system development, stress, robustness, resilience and maturity. Here we attempt to relate to the physical characteristics of the environment, ecosystem indices derived from Ecological Network Analysis. Ten models of food webs were gathered, across a selection of soft-bottom estuaries representative of a large morphological and hydrodynamic diversity, from wave-dominated to mixed energy tide-dominated systems. The selection allowed the comparison of their derived Ecological Network Analysis indices, because of similarities of accuracy in the representation of detritus and bacteria, and because models took into account all trophic levels up to top-predators. In order to obtain comparable physical characteristics, global models were used for a homogeneous description of tide and tidal prisms. Spearman correlations, hierarchical ascendant clustering and Redundancy Analysis were applied to examine the relationship between Ecological Network Analysis indices and physical characteristics. The set of four physical variables selected (catchment area, tidal range at neap tide, index of tide-wave domination and latitude in absolute value) explained 67% of the structure of the Ecological Network Analysis indices. This implies that the physical forcing related to climate, hydrodynamics and morphology is essential for determining the ecological emergent properties of the food web. In the European policy context of determining the ‘good ecological status’ of coastal ecosystems, it implies that the use of Ecological Network Analysis indices for basing the determination of operational indicators should be done, taking into account this context of a strong influence of physical factors.