The material includes 105 children treated for severe thermal injuries in Division of Plastic and Reconstructive Surgery and Burns, University Children’s Hospital of Cracow (Chair and Department of Pediatric Surgery, Polish-American Institute of Pediatrics, Collegium Medicum, Jagiellonian University) and Pediatric Burn Center in the years 1996-2005. The thesis is a retrospective, descriptive and comparative analysis. The data of patients treated in the above-mentioned period were analyzed from the epidemiological and clinical viewpoint. The author gave a detailed description of the dynamic process of development, spread and increase of complexity in colonization and infections of burn wounds in the experimental group, as well as the dependence of microbiological flora on the age of the patient, body surface area and depth involved by the burn, the time interval after the injury and therapeutic modalities. The analysis was performed in 1327 microbacteriological cultures, including 457 cultures of material originating from burn wounds, 429 blood cultures, 215 urine cultures and 226 cultures of material from the endotracheal tubes. The statistical analysis was based on descriptive statistical functions. Qualitative or quantitative data grouped into classes were compared employing χ2–based tests (the χ2 and its modifications – the Fisher’s exact test and the χ2 test with Yates’ mod ; ification), as well as the Kendall-tau coefficient. Two-sided tests were used. Having performed the tests of distribution normality for continuous data, the author employed methods of non-parametric statistical analysis. The significance of inter-group differences was verified using the Mann-Whitney or Kruskal-Wallis tests, respectively. The Spearman’s rank correlation coefficient was used in testing the strength of relationships between particular variables. While comparing the incidence of microorganisms, the author also employed tests serving for comparing two or more structural indicators. Statistical significance was established at p<0.05. The hypotheses were verified at 0.05 (significant), 0.01 (very significant) and 0.001 (highly significant). Based on the results, the author concluded that in the years 1996-2005, the most common strains observed in burn wounds in children included MSCNS, Enterococcus faecalis, MRCNS, Pseudomonas aeruginosa and Acinetobacter baumannii. Positive mycological cultures accounted for 5.8% of all positive microbiological cultures of materials originating from burn wounds. Yeast-like fungi were predominantly isolated, including Candida spp. With a growing time interval after the injury, the microbiological profile of the burn wound changed in the following way: in the first two days, the predominant strains included MSCNS, MSSA and Enterococcus ; faecalis ; between 3-7 days - MRCNS and Enterococcus faecalis ; between 8-14 days - MSCNS, Enterococcus faecalis, MRCNS and Pseudomonas aeruginosa ; between 15-21 days - MSCNS, Enterococcus faecalis, Acinetobacter baumannii and Pseudomonas aeruginosa, and after more than 22 days after thermal injury, the most commonly isolated strains included MSCNS, Corynebacterium sp., MRCNS, Pseudomonas aeruginosa and Acinetobacter baumannii. Yeast-like fungi were present in the burn wound starting in the second half of the first week of burn disease, and starting from the fourth week, they predominated in mycological cultures. Mould fungi were noted in week 2 and 3 following the injury. Treatment protocols did not affect the relative frequency of isolating the majority of bacteria. The exception was Acinetobacter baumannii strains, which were present solely in patients treated by early (day 1-7) burn wound excision followed by grafts (p<0.05), as well as Pseudomonas aeruginosa, which was seen more frequently (p<0.05) in patients treated according to the above protocol. ; It should be borne in mind that the therapeutic protocol was not an independent variable in the study and the selection of treatment protocol was not random, but dictated by the clinical situation, experience of the attending physician and other authors. No statistically significant (p>0.05) positive correlation was demons ; trated between the presence of the same species of microorganisms (or groups of species with similar properties) in the same sub-period of burn disease in the burn wound and blood, burn wound and endotracheal tube and burn wound and urine. The bacterial and fungal species that had been cultured from burn wounds varied from these cultured at the same time from materials originating from other sites. A clear risk factor of bacteremia was the size of the burned body surface area (p<0.001); on the other hand, the depth of the burn was not an independent variable associated with bacteremia. The author demonstrated a correlation between sepsis and body surface area and depth involved by the burn (p<0.001). There was no significant correlation between empiric usage of antibiotics (especially as prevention) and survival rates, development of bacteremia or graft taking (p>0.05). A correlation was demonstrated between the frequency of isolating fungi from the wounds and endotracheal tubes and the amount of antibacterial agents administered during the entire treatment (p<0.001) – more antibacterial drugs led to a later occurrence of fungi, but in such situations, their appearance was more likely. No clinical efficacy of the Pseudovac vaccine was demonstrated in the study group, as vaccination did not affect the incidence of burn wound, blood, urine and endotracheal tube infections by Pseu ; domonas aeruginosa (p>0.05). The arrived at microbiological profile might in the future serve as the basis for developing an algorithm of local and systemic antimicrobial treatment in severe burns in children, as well as basic recommendations for therapy modification depending on individual changes in the “microbiological presentation” of the patients and a possible attempt at formulating the so-called model of microbiological burn wound changeability in children.