A survey was undertaken in the five largest European countries in 1990/91 to measure the incidence of infectious diseases and record their antibiotic treatment, and usage as prophylaxis in acute general hospitals with over 300 beds. A representative sample of between 1234 and 1954 patient records were received for each country, with each participating clinician providing eight to ten records. Data were projected up to national totals. There were similarities between the patients across the countries, 50% of patients received surgery at some time in their hospital stay and in 45% the first antibiotic usage was as prophylaxis. In one fifth of the 55% treated for an infection, that infection was hospital acquired, but these 20% of patients accounted for 30% of total days in hospital for patients treated for infection. Infections of the lower respiratory tract, urinary tract and abdomen accounted for 75% of all infection cases. In 85% of infection cases, initial treatment was empirical. Overall, no culture and sensitivity data were reported by the end of treatment in over 50% of patients, with appreciable variation between the countries. In the UK, 60% of in-patient treatment was with oral antibiotics; in Italy, over 80% was injectable with over half that intramuscular. Germany had the highest rate of intravenous administration at nearly 60% of treatment days. Combination usage was highest, with over 30% of patient treatment days, in France and Spain, but Germany and Italy had the lowest usage at 21% and 16%. Mean duration of administration of antibiotics was shortest in the UK, eight days for infection and four days for prophylaxis, with France the longest at 12 days for infection, and Italy for prophyalxis at over five days. The UK had the highest rate of discharge on antibiotic treatment at over 30% of patients. Initial therapy was effective in about 90% of patients, but the UK, with the shortest durations and the greatest use of older antibiotics often as monotherapy and administered orally, had the highest rate of initial failure (> 10%) of both treatment of infection and prophylaxis. The result of initial failure was a significant prolongation of both hospital stay and antibiotic treatment. The 10% of initial treatment failures accounted for 15% of antibiotic usage in the UK. In France initial treatment failures were treated for twice the length of those in whom the initial therapy was successful.(ABSTRACT TRUNCATED AT 400 WORDS)