Since its discovery at the end of the nineteenth century, Moraxella (Branhamella) catarrhalis has undergone several changes of nomenclature and periodic changes in its perceived status as either a commensal or a pathogen. Molecular analysis based on DNA hybridisation or 16S rDNA sequence comparisons has established its phylogenetic position as a member of the Moraxellaceae and shown that it is related more closely to Acinetobacter spp. than to the genus Neisseria in which it was placed formerly. However, confusion with phenotypically similar Neisseria spp. can occur in the routine diagnostic laboratory if appropriate identification tests are not performed. M. catarrhalis is now accepted as the third commonest pathogen of the respiratory tract after Streptococcus pneumoniae and Haemophilus influenzae. It is a significant cause of otitis media and sinusitis in children and of lower respiratory tract infections in adults, especially those with underlying chest disease. Nosocomial spread of infection, especially within respiratory wards, has been reported. Invasive infection is uncommon, but analysis of reports for England and Wales between 1992 and 1995 revealed 89 cases of M. catarrhalis bacteraemia, with the peak incidence in children aged 1-2 years. Carriage rates of M. catarrhalis are high in children and in the elderly, but its role as a commensal organism has probably been overstated in the past. Approximately 90% of strains are now beta-lactamase positive and, given that the first such strain was reported in 1976, this represents a dramatic increase in frequency over the last 20 years which has not been paralleled in any other species. The BRO-1 and BRO-2 beta-lactamase enzymes of M. catarrhalis are found in other Moraxellaceae, but are not related to beta-lactamases of any other species and their origin is therefore unknown. Molecular and typing studies have shown that the M. catarrhalis species is genetically heterogeneous and these methods have aided epidemiological investigation. Studies of factors that may be related to pathogenicity have shown the existence of three serotypes of lipooligosaccharide and the presence of fimbriae and a possible capsule. Some strains are serum-resistant, probably by virtue of interference with complement action, whilst transferrin- and lactoferrin-binding proteins enable the organism to obtain iron from its environment. An antibody response in humans to various M. catarrhalis antigens, including highly conserved outer-membrane proteins, has been demonstrated. Increased understanding of the organism's pathogenic properties and the host response to it may help to identify suitable vaccine targets or lead to other strategies to prevent infection. Whilst it remains, at present, the third most important respiratory pathogen, the impact of immunisation strategies for other organisms may change this position. The speed with which M. catarrhalis acquired beta-lactamase demonstrates the capacity of this organism to surprise us.