Chronic Pseudomonas aeruginosa lung infections are found in patients suffering from bronchiectasis and cystic fibrosis (CF) and once colonisation is established, it is difficult to remove by current methods. We previously identified a subset of patients with bronchiectasis and chronic P. aeruginosa infection that produced specific antibody that actually protected the infecting bacterium from both serum- and cell-mediated killing. We demonstrated that these ‘inhibitory antibodies’ belong to the IgG2 subclass and target the O-antigen portion of lipopolysaccharide. Crucially, patients with high titres of inhibitory antibodies had worse lung function than infected patients with normal serum killing Two critically ill patients with this IgG2 were treated with plasmapheresis in an attempt to remove the inhibitory antibody. Both patients had immediate benefit from this treatment with a significant drop in hospitalisations, antibiotic use and markers of inflammation. Both patients lost culturable P. aeruginosa in their sputum for up to four months after treatment. Return of the inhibitory antibody in patients coincided with bacteria in their sputum and degrading health. Finally, we investigated the prevalence of this inhibitory antibody in a cohort of CF patients with chronic P. aeruginosa infection. We found that over 34% of the patients had antibody that inhibited serum-mediated killing of P. aeruginosa. Interestingly, titre alone was not sufficient to predict serum-inhibition, with affinity of the antibodies also important. Additionally, we identified some patients with high titres of IgA specific for the O-antigen which was also found to inhibit serum mediated killing. These findings indicate that inhibitory anti-O-antigen antibody may be a significant problem in P. aeruginosa lung infections, and that finding ways to remove or counteract this antibody can lead to improvement in health.