This study sought to investigate the effect of delayed pulsed-wave ultrasound with low frequency on drug release from and the antimicrobial efficacy of vancomycin-loaded acrylic bone cement in vivo and the possible mechanism of this effect. After the implantation of cement and the inoculation of Staphylococcus aureus into the bilateral hips of rabbits, ultrasound (average intensity, 300 mW/cm(2); frequency, 46.5 kHz; on/off ratio, 20 min/10 min) was applied to animals in the normal ultrasound group (UG(0-12)) from 0 through 12 h after surgery and to those in the delayed-ultrasound group (UG(12-24)) from 12 through 24 h after surgery. The control group (CG) was not exposed to ultrasound. Based on vancomycin concentrations in left hip cavities at projected time intervals, the amount of time during which the local drug concentration exceeded the MIC (T(>MIC)) in UG(12-24) was significantly prolonged compared with that in either CG or UG(0-12), and the ratios between the areas under the concentration-time curves over 24 h and the MIC for UG(0-12) and UG(12-24) were both increased compared with that for CG. The greatest reductions in bacterial densities in both right hip aspirates and right femoral tissues at 48 h were achieved with UG(12-24). Local hemorrhage in rabbits of UG(0-12) during the 12-h insonation was more severe than that in rabbits of UG(12-24). Of four variables, the T(>MIC) and the bioacoustic effect were both identified as parameters predictive of the enhancement of the antimicrobial efficacy of cement by ultrasound. Sustained concentrations above the MIC replaced early high maximum concentrations and long-term subtherapeutic release of the drug, provided that ultrasound was not applied until local hemorrhage was relieved. The enhancement of the antimicrobial efficacy of cement by ultrasound may be attributed to the prolonged T(>MIC) and the bioacoustic effect caused by ultrasound.