Anthropogenic stressors can change the associations between marine organisms and their environment. This project uses multivariate statistics and spatial distribution models to assess associations of temperate marine fish assemblages with anthropogenic, environmental, and habitat variables. Fish assemblage data were collected using >1000 BRUVS deployments across Victoria, Australia. To assess anthropogenic impact, we generated proxy variables from existing spatial information (e.g., distance to coast, distance to port, etc.) and combined those with bathymetric structure, SST, wave energy and habitat data observed from the BRUVS. The best DISTLM explained 33.4 % of the variation in fish assemblage across the state and used 24 out of 26 variables including all the anthropogenic impact proxy variables. Overall, depth explained the most variation followed by SST, presence of fine sand, biounit, and wave energy. The spatial models (GAMs) found depth, wave energy and seafloor complexity significant across all models fitted to each of five key species identified from the DISTLMs and for total species richness. Our results show that Victorian fish species and assemblages are strongly influenced by a combination of habitat, environment, and human impact variables, providing valuable information for management of fish communities off Victoria.