We evaluated antimicrobial use in our hospital by department, including indications for use, source of infections, use of the microbiology laboratory, and appropriateness of prescribing, in a prospective, comparative, non-interventional study of all patients receiving antimicrobial agents. We excluded departments where antimicrobial use was negligible. The other 19 departments were followed for 3 (n=4) or 4 (n=15) months, including 2 consecutive months in the spring-summer and either 1 or 2 in the autumn-winter. Antimicrobial therapy was followed from initiation, through possible adaptations, and possible change from intravenous to oral therapy, until discontinuation of treatment. Overall, 6376 antibiotics were given to 2306 patients. Of the surveyed hospitalized patients, 62%+/-22% received antibiotics, with a range of 4-100% per department. Antibiotics were prescribed for infections acquired in the community (3037 instances, 47%), in the hospital (2182, 34%), in a nursing home (575, 9%), and for prophylaxis continued post-operatively (582, 9%). The most common indications for antimicrobial use were: respiratory tract infection (1729, 27%), urinary tract infection (955, 15%), sepsis (701, 11%), intra-abdominal infections (663, 10%), prophylaxis 582 (9%), soft-tissue infection (572, 9%), and surgical site infection (319, 5%). Univariate indicators for appropriateness of treatment were: age, department, site of infection, source of infection, antimicrobial drug and serum creatinine (all p<0.001). Forty-nine antimicrobials were prescribed in 279 combinations, 58% as single agent and 42% as drug combinations. Half of all antimicrobial use consisted of four agents: cefuroxime (19.1%), metronidazole (11.3%), gentamicin (10.6%) and ampicillin (10.2%), which together accounted for 20% of expenditure on antibiotics. Although use of as many as 53% of antimicrobials (26/49) surveyed was restricted, use in this category accounted for only 29% of all antimicrobial courses. Of 6376 antibiotic courses, 4101 (64%) were given intravenously and 2275 (36%) orally. Appropriateness of use of restricted drugs was lower (70%) than of unrestricted ones (84%, p<0.001). Of 24571 defined daily doses (DDD) given orally, 4587 (19%) were restricted, compared to 7264 (34%) of 21602 DDDs given intravenously (p<0.001). Antibiotic treatment in our hospital appears to be substantial and increasing, justifying efforts to improve appropriateness of therapy and improve clinical and financial results.