Twenty-eight patients with large burn injuries (mean total body surface area burned, 47.1%) who underwent 112 wound cleaning and staged early excision procedures were studied prospectively for bacteremia and fungemia induced by wound manipulation. The patients were given an aminoglycoside preoperatively, perioperatively, and postoperatively. Blood samples were obtained immediately before removal of dressings, after wound cleaning, after 30 minutes of surgery, at the end of each procedure, and 1 hour after surgery was completed. In a group of control patients blood samples were obtained immediately before the dressings were removed, after wound cleaning was completed, and 1 hour afterward. In a second group of control patients blood samples were obtained serially while the burn wounds remained undisturbed. None of the control patients received prophylactic aminoglycosides. Induced bacteremia or fungemia was documented in 50 instances of burn wound cleaning and excision; 31 cases of bacteremia or fungemia occurred after wound cleaning alone. Spontaneous bacteremia, i.e., that occurring in the absence of burn wound manipulation, was demonstrated in 3 of 18 blood culture series, whereas induced bacteremia was observed after 11 of the 17 burn wound cleaning procedures alone. The frequency of bacteremia tended to be higher for patients with inhalation injury than for those with no lung involvement. However, lung infection did not seem to account for many of the positive results of blood culture in this study. In contrast, burn wound infection contributed significantly to both spontaneous and induced bacteremia or fungemia. The micro-organisms most frequently isolated were staphylococci (50%) and enterococci (15.1%). Except for three isolates, all micro-organisms were cultured from the burn wounds within 48 hours before each procedure. The induced bacteremia and fungemia were in most cases short lived, and 1 hour after surgery only 9 of 50 blood culture procedures produced positive results. Following three of these procedures, one patient developed pneumonia, whereas two experienced a brief temperature rise, rigors, and slight hypotension without manifesting postoperative sepsis. In the control group undergoing wound cleaning alone, positive blood culture results were demonstrated in 5 of the 11 instances of induced bacteremia. None of these patients showed clinical signs of sepsis during the first 12 hours after wound cleaning. Thus wound cleaning and staged early excision in patients with large burns induced a high frequency of bacteremia or fungemia. The bacteremia and fungemia were transient and caused no serious postoperative infections.