Introduction: Local antimicrobial delivery via calcium sulfate (CaSO 4 ) beads is used as an adjunctive treatment for periprosthetic joint infection. There is limited clinical information describing the performance of antimicrobial-loaded CaSO 4 (ALCS) in large-scale applications. We developed a simulated large joint model to study properties of eluting ALCS. Methods: The in vitro testing platform was an adapted standardized model for tribological testing of prosthetic total hips and total knees (ASTM F732). The model was 70 mL total fluid volume, 25 % bovine serum, and 75 % phosphate-buffered saline, using ISO standard 14242-1 for human synovial fluid simulation. Four brands of CaSO 4 were evaluated. Each 10 mL of CaSO 4 was loaded with 1.2 grams (g) of tobramycin and 1 g of vancomycin powders. A 35 mL bead volume, equaling 175 beads, of each product was placed in incubated flasks. The test period was 6 weeks with scheduled interval fluid exchanges. Fluid samples were tested for antibiotic and calcium concentrations and pH. Results: Antibiotic elution showed an initial burst on Day 1, followed by a logarithmic reduction over 1 week. Tobramycin fully eluted within 2.5 weeks. Vancomycin showed sustained release over 6 weeks. Calcium ion concentrations were high, with gradual decrease after 3 weeks. All four CaSO 4 products were inherently acidic. Fluid became more acidic with the addition of antibiotics primarily driven by vancomycin. Discussion: Clinicians should be cognizant of tobramycin elution burst with ALCS in large loads. The main driver of acidic pH levels was vancomycin. We propose that joint complications may result from lowered fluid acidity, and we suggest clinical study of synovial pH.