OBJECTIVE: To analyse trends in antibiotic use in Dutch hospitals over the period 1997 to 2002. METHODS: Data on the use of antibiotics and hospital resource indicators were obtained by distributing a questionnaire to all Dutch hospital pharmacies. Antibiotic use was expressed as the number of defined daily doses (DDD) per 100 patient-days and as DDD per 100 admissions. RESULTS: Between 1997 and 2002, the mean length of stay decreased by 18%. The mean number of admissions remained almost constant. Total antibiotic use significantly increased by 24%, from 47.2 in 1997 to 58.5 DDD per 100 patient-days in 2002 (p<0.01), whereas expressed as DDD per admissions it remained constant. Antibiotic use varied greatly between the hospitals. Moreover, the mean number of DDD per hospital of amoxicillin with clavulanic acid, clarithromycin, cefazolin, clindamycin and ciprofloxacin increased by 16, 38, 39, 50 and 52%, respectively. Total antibiotic use was higher in university hospitals than in general hospitals. CONCLUSIONS: Between 1997 and 2002, patients hospitalised in the Netherlands did not receive more antibiotics but, since they remained in the hospital for fewer days, the number of DDD per 100 patient-days increased. For macrolides, lincosamides and fluoroquinolones increases in both DDD per 100 patient-days and in DDD per 100 admissions were observed. It is arguable whether these trends result in an increase in selection pressure towards resistance in the hospitals. Continuous surveillance of antibiotic use and resistance is warranted to maintain efficacy and safety of antibiotic treatment.