The number of built structures on the seabed, such as shipwrecks, energy platforms, and pipelines, is increasing in coastal and offshore regions. These structures, typically composed of steel or wood, are substrates for microbial attachment and biofilm formation. The success of biofilm growth depends on substrate characteristics and local environmental conditions, though it is unclear which feature is dominant in shaping biofilm microbiomes. The goal of this study was to understand the substrate- and site-specific impacts of built structures on short-term biofilm composition and functional potential. Seafloor experiments were conducted wherein steel and wood surfaces were deployed for four months at distances extending up to 115 m away from three historic (>50 years old) shipwrecks in the Gulf of Mexico. DNA from biofilms on the steel and wood was extracted, and metagenomes were sequenced on an Illumina NextSeq. A bioinformatics analysis revealed that the taxonomic composition was significantly different between substrates and sites, with substrate being the primary determining factor. Regardless of site, the steel biofilms had a higher abundance of genes related to biofilm formation, and sulfur, iron, and nitrogen cycling, while the wood biofilms showed a higher abundance of manganese cycling and methanol oxidation genes. This study demonstrates how substrate composition shapes biofilm microbiomes and suggests that marine biofilms may contribute to nutrient cycling at depth. Analyzing the marine biofilm microbiome provides insight into the ecological impact of anthropogenic structures on the seabed.