Objective Platelet-rich plasma (PRP) contains high concentrations of platelets and leucocytes, which play a key role in antimicrobial host defense system. To evaluate the antimicrobial efficacy of autologous PRP in vitro and in vivo and to explore the mechanism of action so as to provide the experimental basis for the prevention and treatment of bone infection. Methods PRP was prepared with the method of two centrifugation from 15 health volunteers. Platelet-leukocytegel (PLG) was obtained after activation of PRP with bovine thrombin. Next, PLG was incubated with Staphylococcus aureus (1 × 106 cfu/mL) in vitro compared with PRP, platelet-poor plasma (PPP) and PBS. Samples were taken out after 2, 4, 6, 8, 12, and 24 hours for bacterial culture and colony count. Thirty-six New Zealand adult rabbits, weighing (2.85 ± 0.11) kg, were divided into 4 groups: PLG (n=10), antibiotic (n=10), infection (n=10), and PBS (n=6) groups. The osteomyel itis models were made by injecting 0.1 mL Staphylococcus aureus suspension (1 × 106 cfu/mL) into the tibial canal in PLG group, antibiotic group, and infection group; equal volumes of PBS was injected in PBS group as a control. Autologous PLG was injected immediately after operation in PLG group. Cefazol in (30 mg/kg) was injected through the auricular vein from 1 hour before operation to 72 hours after operation in antibiotic group, once per 8 hours. No treatment was given in infection and PBS groups. The efficacy of PLG for osteomyel itis prophylaxis was evaluated by microbiological, X-ray and histological observation within 28 days. Results The contents of leucocyte and platelet of PRP were 6.2 times and 5.5 times of whole blood, showing signficant differences ((P lt; 0.05); the contents of leucocyte and platelet of PPP were significantly lower than those of whole blood and PRP ((P lt; 0.05). In vitro test showed that PLG had the most obvious bacteriostasis effect. The bacterial count reached a minimum value at 4 hours after incubation in PLG and at 6 hours after incubation in PRP. PPP had slow and no obvious bacteriostasis effect and PBS had no bacteriostasis effect. At 2, 4, 6, 8, 12, and 24 hours of incubation, the bacterial count reduced significantly when compared PLG with PRP and PPP (P lt; 0.05), when compared PRP with PPP (P lt; 0.05). In PLG group and antibiotic group, 1 rabbit died, respectively; 34 rabbits survived to the end of the experiment. There was no significant difference (P gt; 0.05) in temperature, body weight, erythrocyte sedimentation rate and content of leucocyte between 28 days after operation andbefore operation in 4 groups. After 28 days, the X-ray scores were 2.78 ± 1.39, 1.55 ± 1.48, 4.17 ± 1.25, and 0 in PLG, antibiotic,infection, and PBS groups, respectively, which was significantly higher in infection group than in other 3 groups ((P lt; 0.05). Also, the histological scores were 5.89 ± 3.92, 3.00 ± 2.31, 10.33 ± 4.03, and 0, respectively, which was significantly higher in infection group than in other 3 groups (P lt; 0.05), and was significantly lower in antibiotic group than in PLG group ((P lt; 0.05). The results of bacterial culture showed that the infection rates of PLG group (44.4%) and antibiotic group (20.0%) were significantly lower ((P lt; 0.05) than that of infection group (88.9%). The quantitative analysis of bacteria showed that the number of bacteria was signifcantly lower ((P lt; 0.05) in PLG and antibiotic groups than in infection group. Conclusion PRP forms into PLG after activating, it can inhibit Staphylococcus aureus reproduction in vitro and can effectively prevent bone infection in vivo.