To investigate the bacterial epidemiology of blood cultures taken during the treatment of critically ill burn patients, the use of antibiotics at admission and before the observation of positive blood cultures, and their effect on prognosis. A retrospective study method was used. From January 1, 2010, to December 31, 2019, burn patients who met the inclusion criteria and were treated at the Burn Department, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, were enrolled in the study. Data were collected from the patients’ electronic medical records. General patient information, including length of hospital stay, length of intensive care unit (ICU) stay, in-hospital mortality, the bacteria epidemiological characteristics of blood cultures, and the use of antibiotics within 24 h after admission and before a positive blood culture was observed, was collected. Independent sample t tests and χ2 tests were used to compare the effects of a positive blood culture and the use of appropriate antibiotics within 24 h after admission and before the observation of a positive blood culture on prognosis. (1) The three most frequently detected bacteria in the blood cultures were Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii, and the amount of K. pneumoniae detected increased gradually. (2) Compared with the group of patients who were negative for bloodstream infection , the positive bloodstream infection group had a larger total body burn surface area (TBSA) (t = − 5.097, P = 0.000) and third-degree burn area (t = − 5.133, P = 0.000), a significantly longer length of hospital stay (t = 3.003, P = 0.003) and the length of ICU stay (t = 4.258, P = 0.000), and a significantly higher rate of in-hospital mortality (χ2 = 8.485, P = 0.004). When K. pneumoniae was detected, the length of hospital stay (t = 2.148, P = 0.035) and the length of ICU stay (t = 2.880, P = 0.005) were significantly prolonged. (3) The two antibiotics that were most frequently used in patients with acute burns within 24 h after admission were lincomycin (90 cases, 29.32%) and carbapenems (79 cases, 25.73%). Comparing the clinical characteristics of the lincomycin group and the carbapenem group, the TBSA (t = − 3.34, P = 0.001) and the third-degree burn area (t = − 6.08, P = 0.000) of the patients in the carbapenem group were larger, and the length of hospital stay (t = − 2.136, P = 0.035) and length of ICU stay (t = − 5.18, P = 0.000) were longer, but the difference in in-hospital mortality was not statistically significant (χ2 = 1.983, P = 0.159). (4) Comparing the group with appropriate initial antibiotic use within 24 h of admission to the inappropriate use group, the TBSA (t = − 0.605, P = 0.547), the third-degree burn area (t = 0.348, P = 0.729), the length of hospital stay (t = − 0.767, P = 0.445), the length of ICU stay (t = − 0.220, P = 0.827) and in-hospital mortality (χ2 = 1.271, P = 0.260) were not significantly different. (5) Comparing the group with appropriate antibiotic use before a positive blood culture was observed to the group with inappropriate antibiotic use, the TBSA (t = − 0.418, P = 0.677), the third-degree burn area (t = 0.266, P = 0.791), the length of hospital stay, the length of ICU stay (t = 0.995, P = 0.322) and in-hospital mortality (χ2 = 1.274, P = 0.259) were not significantly different. We found that patients with a positive blood culture had a larger burn area and a worse prognosis; that the greater the amount of K. pneumoniae in the bloodstream of burn patients was, the longer the hospital and ICU stays were; that whether appropriate antibiotics were administered to acute critical burn patients 24 h after admission had no effect on the prognosis; and that whether appropriate antibiotics were administered before a positive blood culture was observed had no effect on prognosis.