Traditionally, organisms responsible for major biogeochemical cycling processes have been determined by physiological characterization of environmental isolates in laboratory culture. Molecular techniques have, however, confirmed the widespread occurrence of abundant bacterial and archaeal groups with no cultivated representative, making it difficult to determine their ecosystem function. Until recently, ammonia oxidation, the first step in the globally important process of nitrification, was thought to be performed almost exclusively by bacteria. Metagenome studies, followed by laboratory isolation, then demonstrated the potential for significant ammonia oxidation by mesophilic crenarchaea, whose ecosystem function was previously unknown. Re-assessment of the role of bacteria in ammonia oxidation is now required and this article reviews the current evidence for the relative importance of bacteria and archaea. Much of this evidence is based on metagenomic analysis and molecular techniques for estimation of gene and gene transcript abundance, changes in ammonia oxidizer community structure during active nitrification and phylogeny of natural communities. These studies have been complemented by physiological characterization of a laboratory isolate and by incorporation of labelled substrates. Data from these studies provide increasingly convincing evidence for the importance of archaeal ammonia oxidizers in the global nitrogen cycle. They also highlight the need to re-assess the importance of ammonia-oxidizing bacteria, the requirement and limitations of molecular techniques in linking specific microbial groups to ecosystem function and the limitations of reliance on laboratory cultures.