Sixteen patients with chronic bronchitis were treated in a single-blind crossover study during two separate acute exacerbations caused by bacterial infections. During one of the episodes a regimen consisting of two 800-mg doses of bacampicillin per day was given; therapy for the other episode was four 500-mg doses of ampicillin per day. Each drug was given for 14 days. Patients were observed once before therapy was started, twice a week during therapy, and weekly after therapy; the investigator did not know which drug the patients were receiving. Graded clinical observations included frequency and severity of cough, wheezing, rales, rhonchi, breath sounds, and prolongation of expiration. Objective sputum measurements included daily volume, qualitative and quantitative cytology, and quantitative bacterial counts. Paired t-test analyses revealed that before therapy was begun the group receiving bacampicillin was sicker. Both agents effectively decreased inflammation and counts of bacterial flora. Clinical symptoms cleared 11 days after institution of therapy with either drug, but patients tended to favor bacampicillin in overall clinical effects. Adverse effects were similar with both agents, but bacampicillin was associated with fewer upper and more lower gastrointestinal symptoms.