The CO2 laser prevents bleeding by sealing blood and lymph vessels as it vaporizes tissue. Bacteremia following oral surgery might not occur under these conditions. To test this hypothesis, a 0.2-mm-deep incision 1 cm long was made in the right buccal cheek pouch of hamsters using either laser, electrosurgery, or a scalpel. Twenty minutes later, 1 mL of blood was taken from each animal by cardiac puncture, inoculated on a blood agar medium, and incubated anaerobically for 4 days; then the colonies were counted. Using an operational definition of bacteremia as five colonies or more per plate, there were no positive results out of 18 trials (0/18) for laser surgery, 7/8 for electrosurgery, and 8/12 for scalpel surgery. Based on the Student t test using the binomial distribution, the laser produced statistically less bacteremia than the other two methods (P less than .01). Because the five-colony cutoff was arbitrary, the nonparametric Wilcoxon Rank test was also used. Colony formation from blood from the laser group was significantly less than from the electrosurgery group (P less than .01) and the scalpel group (P less than .05). The laser surgery group was not statistically different from the control (nonsurgerized) group. These results suggest that there is a considerable bacteremia following scalpel and electrosurgery, but that laser surgery produces no bacteremia.