Sandy beaches, forming the most widespread coastal habitat in the world, are threatened by the effects of increasing anthropogenic pressure in the context of global change, including the increased occurrence of green tides composed of free-living Ulva spp. Sandy beaches are also highly dynamic ecosystems that support numerous essential ecological functions and contain a distinctive biodiversity, but their precise functioning and natural variability (i.e. the disentangling of biological and physical influences) remain under-studied. Our study aimed at determining the effects of space, time, and environmental variables on the natural variability of macrofaunal community structure and at specifically determining the effects of macroalgal accumulations on the observed variability. We followed a high-resolution field sampling design in space and time (261 samples) at non-vegetated and at partially and fully green-tide-impacted macrotidal sandy beaches. We used novel statistical approaches (distance-based Moran’s eigenvector maps [dbMEMs] and variation partitioning) to analyse our results. The macrofaunal community structure of the non-vegetated sandy beach was variable in space and time at small scales, and physical environmental variables significantly explained these variations. Our study also highlighted a decrease in this variability along a gradient of increasing coverage of stranded Ulva spp. and the increasing importance of biological variables in explaining ecological variability. Compared to a state with no Ulva, a large and homogeneous coverage of Ulva along a sandy beach shore induced a significant decrease of overall β-diversity. However, macrofaunal responses to macroalgal accumulation were less pronounced than what has been previously shown in micro-tidal and sheltered systems, likely due to the dynamic and unstable nature of high-energy macrotidal sandy shores.