Prosthetic joint infection (PJI) is a rare but devastating complication of joint arthroplasty. Biofilm formation around the prosthesis confers tolerance to antibiotics so that treatment is challenging. Most animal models of PJI use planktonic bacteria to establish the infection which fails to reproduce the pathology of chronic infection. We aimed to establish a rat model of Staphylococcus aureus PJI in male Sprague–Dawley rats using biofilm inocula and demonstrate its tolerance to frontline antibiotics. Pilot studies indicated that infection could be introduced to the knee joint by a biofilm-coated pin but that handling the prosthetic without disturbing the biofilm was difficult. We, therefore, developed a pin with a slotted end and used a miniature-biofilm reactor to develop mature biofilm in this niche. These biofilm-laden pins consistently produced infection of the bone and joint space. Treatment with high dose cefazolin, 250 mg/kg, starting the day of surgery reduced or cleared pin-adherent bioburden within 7 days, however when escalation from 25 to 250 mg/kg cefazolin treatment was delayed for 48 h, rats were unable to clear the infection. To track infections, we used bioluminescent bacteria, however, the bioluminescent signal did not accurately track the degree of infection in the bone and joint space as the signal did not penetrate the bone. In conclusion, we demonstrate that using a custom prosthetic pin, we can generate biofilm in a specific niche using a novel bioreactor setup and initiate a rat PJI that rapidly develops tolerance to supra-clinical doses of cefazolin.