Background: Regular intramuscular (i.m.) benzathine penicillin G (BPG) injections have been the cornerstone of rheumatic heart disease (RHD) secondary prophylaxis since the 1950s. Patient adherence to IM BPG is poor, largely due to pain, the need for regular injections every 3–4 weeks and health sector delivery challenges in resource-limited settings. There is an urgent need for new approaches for secondary prophylaxis, such as an implant which could provide sustained penicillin concentrations for more than 6 months. Methods: In this study we developed and evaluated a slow release implant with potential for substantially extended treatment. The side wall of a solid drug rich core was coated with polycaprolactone which acts as an impermeable barrier. The exposed surfaces at the ends of the implant defined the release surface area, and the in vitro release rate of drug was proportional to the exposed surface area across implants of differing diameter. The in vivo pharmacokinetics and tolerability of the implants were evaluated in a sheep model over 9 weeks after subcutaneous implantation. Results: The absolute release rates obtained for the poorly water-soluble benzathine salt were dependent on the exposed surface area demonstrating the impermeability of the wall of the implant. The implants were well-tolerated after subcutaneous implantation in a sheep model, without adverse effects at the implantation site. Gross structural integrity was maintained over the course of the study, with erosion limited to the dual-exposed ends. Steady release of penicillin G was observed over the 9 weeks and resulted in approximately constant plasma concentrations close to accepted target concentrations. Conclusion: In principle, a long acting BPG implant is feasible as an alternative to i.m. injections for secondary prophylaxis of RHD. However, large implant size is currently a significant impediment to clinical utility and acceptability.