The objective of this study was to determine if three different concentrations of cefazolin and penicillin irrigation solutions reduce quantitative bacterial counts in experimental crush wounds contaminated with multiple species of bacteria. The design used was a randomized, blinded, experimental animal study. An animal bite wound model was created by innoculating crushed incisions with three species of bacteria. Four paravertebral incisions extending to deep fascia were created in each of twelve anesthetized albino guinea pigs. Wound edges were clamped with a hemostat for five seconds to create crushed, devitalized tissue within each wound. Wounds were inoculated with 0.4 mL of a standard solution of Staphylococcus aureus, Bacterioides fragilis, and Pasturella multocida and covered. Four hours after inoculation, each wound was scrubbed for 30 seconds with 20% poloxamer 188 and then irrigated with 100 mL of one of four solutions: normal saline solution (control); cefazolin (CZ) 2 mg/mL, plus penicillin G (PCN) 200 units/mL (low dose); CZ 10 mg/mL, plus PCN 2,000 units/mL (intermediate dose); and CZ 50 mg/mL, plus PCN 20, 000 units/mL (high dose). Investigators were blinded to the solutions used. Wounds were covered with a vapor-permeable dressing. Six days after treatment, each wound was examined for signs of infection and then excised for quantitative bacteriologic analysis. Colony counts were reported as counts per gram of tissue. Wounds in the four irrigation solution groups were compared using ANOVA. A log difference of 3 was considered significant. The average log total bacteria/gram tissue for the four groups were: control, 4.35 (95% CI; 1.01); low dose, 4.09 (95% CI; 1.42); intermediate dose, 4.47 (95% CI; 1.27); and high dose, 3.45 (95% CI; 1.33). No wounds in the high-dose group had any clinical signs of infection, whereas 50% of wounds in the intermediate dose group, 42% in the low dose group, and 33% in the control group had either erythema, induration, or purulence. There were no statistically significant differences in the bacterial counts/gram tissue or clinical infection rates in any of the groups. A formal trend analysis failed to find a significant linear trend for decreasing bacterial counts for either antibiotic. In this experimental bite wound model containing contaminated, crushed tissue, irrigation with various solutions of cefazolin plus penicillin G did not reduce quantitative bacterial counts more than 3.1 log total bacteria/gram tissue.