At the workshop on infections after stem cell transplantation (SCT) in children, the following topics were introduced by invited speakers and discussed with the audience: empirical antimicrobial therapy in the pre-engraftment period, early diagnosis of fungal and viral infections and possibilities to treat them and the possible role of G-CSF early post-SCT. Episodes of fever in the pre-engraftment period mostly are unexplained, and in about one quarter due to bacteremia, mostly by Gram-positive cocci. No single drug or combination of drugs used for antimicrobial therapy is superior, neither does it cover 100% of the pathogens. Close microbiological surveillance of the patients and knowledge of the local microbial epidemiology are requested for optimal therapy. Early fungal infections are reactivations of pre-SCT infections, late fungal infections mostly are associated with failure of engraftment or GvHD and its treatment. Except for suggestive ultrasound or CT-scan abnormalities, the possibilities for early diagnosis are limited c.q. not reliable. Fluconazol prophylaxis is recommended to prevent Candida albicans invasion. A number of new antifungal drugs are being tested in phase I and II studies. CMV, EBV and adenoviruses may reactivate after SCT, causing severe disease with a high mortality, especially in non-HLA-identical donor-recipient combinations. Frequent surveillance cultures for CMV and adenoviruses, pp65-CMV antigen detection in WBC and PCR techniques for CMV, EBV and adenoviruses all have their own contribution to the early diagnosis of dissemination of the viral infection. Therapeutical possibilities, except with respect to ganciclovir and foscarnet for CMV infection, are still limited. The effectiveness of cidofovir is under study. Adoptive therapy with virus-specific CTL's probably represents the new frontier. G-CSF administration early after SCT has a beneficial effect on PMN recovery, hospitalization time, use of antibiotics and total parenteral nutrition requirement in children undergoing allogeneic and autologous BMT. No benefit is observed in children undergoing peripheral blood SCT. The routine use of G-CSF in the latter group of patients is not justified.