This thesis reports the characterisation of a novel outer membrane protein (OMP) from M. catarrhalis, designated M35, with a molecular mass of 36.1 kDa. This protein is structurally homologous to classic Gram-negative porins, such as OMP C from E. coli and OMP K36 from K. pneumoniae, with a predicted structure of 8 surface loops connecting 16 antiparallel -sheets. Comparison of the DNA sequences of the M35 genes from 18 diverse clinical isolates showed that the gene was highly conserved (99.6-100 % of nucleotides) with only one isolate (ID78LN266) having base variations that resulted in amino acid substitutions. A single amino acid mutation in the 3rd external loop of M35 in isolate ID78LN266 significantly affected antibody recognition, indicating that loop 3 contains an immunodominant B-cell epitope. The reduction in antibody-binding to M35 from ID78LN266 was similar to that caused by complete removal of loop 3. Since loop 3 folds into the porin channel in the classic structure, the antibody specificity to loop 3 was hypothesised to be a potential mechanism for evasion of host immune responses targeted to M35, potentially explaining the high degree of conservation across isolates. A series of recombinant proteins were constructed to analyse the binding to M35 of antibodies specificity for loop 3 or the remainder of the protein. It was found that loop 3- specific antibodies were not able to bind to M35 on the surface of M. catarrhalis and that this corresponds both with a lack of ability to enhance opsonophagocytosis in vitro and bacterial clearance in vivo. Additionally, antibodies raised against a version of M35 lacking loop 3 and M35 from the variant isolate ID78LN266 were both no less effective than the full consensus M35 by both these measures. It therefore appears that while the majority of antibodies raised against M35 are specific for loop 3 these antibodies do not mediate anti-M. catarrhalis actions. Two deletion mutant strains of M. catarrhalis that do not contain the outer membrane protein M35 were created by insertional inactivation of the M35 gene. Growth comparisons between these mutant strains and their wildtype parent strains initially led to the hypothesis that M35 is necessary for efficient glutamic acid uptake by M. catarrhalis, however this hypothesis was later shown to be incorrect. Efficient uptake of glutamic acid seemed to be mediated by a novel 40 kDa protein that was up-regulated in the deletion mutant strains, presumably to compensate for the lack of M35. M35 was also found to be essential for in vivo survival of M. catarrhalis in the nasal cavities of mice, indicating that it is an essential functional protein for colonisation of the mucosal surface.