Archean microfossils provide some of the earliest physical evidence for life on Earth, yet there remains a great deal of uncertainty regarding which micro-organisms were actually preserved. Because of the limited cellular detail remaining, interpretation of those microfossils has been based solely on size and morphology. This has led to significant controversy surrounding the presence or absence of cyanobacteria as early as 3.5 billion years. Accordingly, there has been an experimental bias towards studying their silicification. Here we report the very first findings on thermophilic bacteria–silica interactions, and investigate how Sulfurihydrogenibium azorense, a representative of the Aquificales often found as prominent members of modern hot spring vent communities, interacts with highly siliceous hydrothermal fluids. We show that adsorption of silica is limited to silica polymers and colloids, and that the magnitude of silica adsorption is dependent on its chemolithoautotrophic pathway. Intriguingly, when S. azorense is grown as a H2-oxidizer, it responds to increasing silica concentrations by producing a protein-rich biofilm that may afford the cells protection against cell wall silicification. Although the biofilms of Aquificales could potentially contribute to or accelerate siliceous sinter formation under certain growth conditions, the cells themselves show a low preservation potential and are unlikely to have been preserved in the ancient rock record, despite phylogenetic analyses suggesting that they represent one of the most primordial life forms.