The frequency with which bacterial infection causes exacerbations of chronic obstructive pulmonary disease (COPD) may depend on the dominant pathology present; patients with chronic bronchitis are more susceptible to bacterial bronchial infections than those at the emphysema or asthma ends of the spectrum. However, impairment in respiratory function may be very important in governing the outcome of an exacerbation. Placebo-controlled trials have provided conflicting evidence of the efficacy of antibiotics in acute exacerbations. Overall, there is a significant benefit, particularly in certain patient groups, defined by symptoms and past history. Haemophilus influenzae, Streptococcus pneumoniae, and Moraxella catarrhalis are the species most commonly isolated during exacerbations, and the same species may colonize the bronchial mucosa when the patient is in a stable state. Evidence is accumulating that bacteria are an independent stimulus of mucus hypersecretion and bronchial inflammation, and that they interact with other stimuli such as viral infection, atmospheric pollution, and tobacco smoke. New approaches are being used to investigate the importance of bacterial infection in patients with COPD. There are several good reasons why new more potent antibiotics might be expected to be superior to older standard compounds in the management of patients with problematic COPD. However, future studies should aim to confirm that bacteriologic superiority translates into improved clinical outcomes, and seek to measure the level of benefit.