Ascertainment of SBIs plays a central role in their management, which can affect the prognosis, hopefully avoiding an inappropriate antibacterial therapy concerning choice, dosing, timing, duration and route of administration of antibiotics. Different aspects of SBI management were evaluated by interviewing doctors practising in ICU, Surgery and Haematology wards. In the period 16 June - 7 July 2003, 150 doctors, equally distributed by specialty and geographical location, experienced in the management of antibiotic therapy, were interviewed in order to acquire the following information: criteria adopted to define SBIs, presumed incidence, most frequent diagnosis, initial approach to antibiotic therapy (empirical or not, route of administration, mono- or combination therapy), ID consultation request. In most cases generic and empirical criteria are used to define SBI, generally associated to the presence of co-morbidities, the highest rates being reported in ICUs (35.1%) and Haematology (34.7%) wards. Pneumonia is the top reported SBI in all the wards, followed by sepsis in ICUs and Haematology, and by intrabdominal infections in Surgery. Antibiotic therapy is often empirical (~90%), often performed i.v. with antibiotics given in combination. Following treatment failure, which occurs on average in 35.5% of cases, ID consultation and microbiological investigation are required. ID consultation is required in 20.2%, 26.1% and 28.1% of cases by haematologists, surgeons and ICU specialists, respectively. SBIs frequently occur in all the wards where the interviews were conducted. Their management is generally empirical and either ID consultation or microbiological investigation is infrequently required especially as an initial approach. The use of appropriate guidelines and ID consultation, as proven in many controlled studies, could be efficacious in reducing the incidence of inappropriate therapies and increasing favourable outcome rates.