Among the main aspects to be considered when treating burns, the problem of infection control remains unsolved. Considerable financial resources are needed to prevent the transmission of organisms. To justify such investments in buildings and antiseptic measures, an extensive epidemiological hospital study was carried out from 1970 to 1974, involving 930 patients, and more than 25,000 wound biopsies as well as 10,000 contact cultures and environmental swabs. Bacteria from the environment of severly burned patients were counted every week. Serotyping was used for a specialized study of Pseudomonas aeruginosa. In 200 patients wound organisms were counted. The most important organisms were: Streptococcaceae (pyogenic streptococci, less frequently faecal and salivary streptococci). Pseudomonadaceae, Enterobacteriaceae, and Micrococcaceae (especially Micrococcus aureus). Povidon iodine, gentamicin and silver sulfadiazine were used for local disinfection. Antibiotics used were gentamicin, carbenicillin and polymyxin. Whereas from 1970 to 1972 P. aeruginosa was the predominant organism found in wounds, other gram-positive organisms increased from 1972 on. Wounds were colonized mainly in the course of the first two weeks of treatment. Special studies regarding P. aeruginosa revealed a predominance of serotypes 5 and 13 between 1970 and 1973, whereas types brought into the hospital were dominant from 1973 on. An analysis of furniture and equipment, water faucets and drains showed that Pseudomonas strains found in the water did not coincide with those found in wounds. Therefore, a contamination from this source seems unlikely. Strains found on furniture and equipment, however, also appeared in the wound flora. When the therapeutic routine was changed (to prevent patients passing through common treatment areas such as bathrooms and dressing areas) hospital organisms 5 and 13 could be eliminated almost completely. Thus, it is possible to achieve a considerable reduction in the rate of cross-infection among patients by, for instance, excluding common treatment areas from the therapy programme. Nevertheless, in the majority of cases wounds will still be colonized, in particular by bacteria that were already in the anal region or on the skin before the patient was injured. For this reason, the elimination of such organisms by topical bactericidal agents constitutes an an important factor in efforts to reduce the rate of septicaemic complications. In view of the persisting high mortality due to generalized infections this therapeutic aspect must also be exploited thoroughly in the future. Although in comparative studies of topical therapy using povidon iodine, silver sulfadiazine and gentamicin, organisms did appear in the course of the first two weeks; in the case of the PVP-I the colonization never reached 10(5) organisms per cm2, i.e. the danger threshold for generalized sepsis. There was no evidence of a correlation between number of organisms and depth of burns.