The benefit of antimicrobial therapy for patients with an acute exacerbation of chronic bronchitis (AECB) remains controversial for two main reasons. First, the distal airways of patients with chronic bronchitis are persistently colonised, even during clinically stable periods, with the same bacteria that have been associated with AECB. Second, bacterial infection is only one of several causes of AECB. These factors have led to conflicting analyses on the role of bacterial agents and the response to antimicrobial therapy of patients with AECB. An episode of AECB is said to be present when a patient with chronic obstructive pulmonary disease (COPD) experiences some combination of increased dyspnoea, increased sputum volume, increased sputum purulence and worsening lung function. While the average COPD patient experiences 2 - 4 episodes of AECB per year, some patients, particularly those with more severe airway obstruction, are more susceptible to these attacks than others. Bacterial agents appear to be particularly associated with AECB in patients with low lung function and those with frequent episodes accompanied by purulent sputum. Non-typeable Haemophilus influenzae, Streptococcus pneumoniae and Moraxella catarrhalis account for up to 50% of episodes of AECB. Gram-negative bacilli are more likely to occur in patients with more severe lung disease. Antibiotics have been used to ameliorate AECB, to prevent AECB and to prevent the long-term loss of lung function that characterises COPD. Numerous prevention trials have been conducted with fairly consistent results; antibiotics do not lessen the number of episodes of AECB but do reduce the number of days lost from work. Most antibiotic trials have studied the impact of treatment on episodes of AECB and results have been inconsistent, largely due to patient selection and end point definition. In patients with severe airway obstruction, especially in the presence of purulent sputum, antibiotic therapy significantly shortens the duration of symptoms and can be cost-effective. Over the past 50 years, virtually all classes of antimicrobial agents have been studied in AECB. Important considerations include penetration into respiratory secretions, spectrum of activity and antimicrobial resistance. These factors limit the usefulness of drugs such as amoxicillin, erythromycin and trimethoprim-sulfamethoxazole. Extended-spectrum oral cephalosporins, newer macrolides and doxycycline have demonstrated efficacy in clinical trials. Amoxicillin-clavulanate and flouoroquinolones should generally be reserved for patients with more severe disease. A number of investigational agents, including ketolides and newer quinolones, hold promise for treatment of AECB.