Patients with purulent exacerbation of chronic bronchitis were randomized to receive either a single 400-mg daily dose of cefixime or 250 mg of cephalexin, orally, four times a day. Patients were males with a mean age of 63 years. Of the 86 patients, 71 (82%) had bronchitis caused by a single organism (29 by Haemophilus influenzae, 27 by Branhamella catarrhalis, 9 by gram-negative enteric organisms, 6 by Streptococcus pneumoniae), while more than one pathogen was implicated in 15 patients (18%). A total of 70.8% of the cefixime group and 50% of the cephalexin group were clinically cured (chi 2 = 3.89, P less than 0.05); however, when the categories of cured and improved were combined, no significant difference was noted between treatment groups (chi 2 = 3.39, P = 0.06). Analysis of side effects included all 130 evaluable and nonevaluable patients: diarrhea was noted in six patients in the cefixime group and none of the patients in the cephalexin group (P = 0.013 by the Fisher exact test). The diarrhea was mild and self-limited in all cases. B. catarrhalis has emerged as a major cause of exacerbation of bronchitis in our experience; there is an increased need to emphasize the examination of sputum samples by Gram staining if cost-effective antibiotic choices are to be made; any empirically chosen antibiotic should have activity against beta-lactamase-producing strains of B. catarrhalis as well as S. pneumoniae and H. influenzae.