The morphological variability of freshwater bivalve species, observed between and within river basins, may hamper their correct identification, even by experienced researchers. Classic morphometric measurements, i.e. shell length, height and thickness, or their ratios, are generally insufficient to distinguish populations and/or species. These issues may be overcome using a geometric morphometric method, which allows analysis of the overall shape of the individual, independently of its size. Thus, we aimed to test the usefulness of two geometric morphometric tools, landmarks and sliding semi-landmarks, to evaluate the morphological variability of Unio delphinus Spengler, 1783 in three habitats of the Guadiana Basin (SW Iberian Peninsula, Europe): estuary, river and stream. We used 13 landmarks located on the shell interior (at the teeth, muscle scars and pallial line) and 35 sliding semi-landmarks for the shell contour. These morphometric analyses showed that the shell shape of U. delphinus is differs significantly among different habitats. Estuarine and stream shells are the most disparate (James Index= 649.114, permutation p-value< 0.001), and variability is not related to variations in shell size. The main differences in shell morphology are the following: a) estuarine shells are more elongate, while riverine shells are more subovate; b) the anterior curvature at the umbo is steeper in estuarine and riverine shells; c) estuarine shells have an arched curvature at the ventral part of the shell, which is absent in specimens from the other habitats. Our data suggests that the morphology of U. delphinus shells might be influenced by the water flow characteristics of each habitat, since shells exhibited characteristics that are typically observed in freshwater mussels from lotic and lentic habitats.