A double-blind prospective study was done to assess the benefit of delaying closure of the wounds associated with open fractures. An additional doubleblind study compared the effectiveness of clindamycin versus cefazolin for prophylactic antibiotic coverage. Quantitative cultures of the wounds were accomplished at the time of debridement and again at the time of closure if the wound was not closed initially.Almost half of the wounds were contaminated (46%) at the time of debridement, although the incidence of wound infection was low (6.5%).Gram-negative organisms resistant to the prophylactic antibiotic were recovered initially only eight times, but four of these (50%) became infected. The contaminating organisms in each case were present in high concentration (greater than 105CFU/gm of tissue) at initial culture. The time of wound closure, cefazolin versus clindamycin, and internal fixation of the fracture were not followed by significant differences in the development of clinical infection in this series.