92 patients with antibiotic-associated diarrhea were randomized to receive oral teicoplanin 100 mg twice daily for 7 days (BID group); or 50 mg 4 times daily for 3 days, followed by 100 mg twice daily for 4 days (QID group) in a randomized, double-blind, multicentre study. Clostridium difficile was demonstrated by culture and/or cytotoxin test in 49 (53%) patients, of whom 47 (23 male, 24 females, mean age 65 years; 23 in the BID group, 24 in the QID group) were evaluable for clinical efficacy. Prior treatment with cephalosporins was registered in 49%, isoxazolyl-penicillins in 33% and clindamycin in 20% of the C. difficile positive patients. On the last day of treatment, 96% (23 of 24 patients) in the QID group were found cured, compared with 70% (16 of 23 patients) in the BID group (p = 0.02). On days 2 and 3 of treatment, QID group patients had significantly fewer loose stools per day (p < 0.05) than those of the BID group. Clinical recurrence, within 4 weeks post-treatment, occurred in 35% and 33% of the patients in the BID and QID groups, respectively. The bacteriological elimination rate 4 weeks post-treatment was 55% in the BID group and 59% in the QID group. The study was terminated prematurely due to the unexpectedly high clinical failure and recurrence rate in C. difficile positive patients treated with the BID dosage regimen.