The current prophylactic treatment to prevent rheumatic heart disease requires four-weekly intramuscular injection of a suspension of the poorly soluble benzathine salt form of penicillin G (BPG) often for more than 10 years. In seeking to reduce the frequency of administration to improve adherence, biodegradable polymer matrices have been investigated. Poly(lactide-co-glycolide) (PLGA)-based in situ forming precursor systems containing N-methyl-2-pyrrolidone as solvent and PLGA-based monolithic implants for surgical implantation containing BPG were developed. Long-term release studies indicated low and plateaued release of penicillin G, but continual favourable release profiles for the benzathine counterion, indicating degradation of the polymer and generation of acidic microenvironment being detrimental to penicillin stability. In order to avoid the issue of the acidic product, poly(caprolactone)(PCL) implants were also investigated, with favourable penicillin G release behaviour being achieved, and slow release over 180 days. However, when taking into account the mass of polymer, and the total dose of drug calculated from literature pharmacokinetic parameters for penicillin G, we concluded that an implant size of over 7 g would still be required. This may preclude clinical deployment of a polymer matrix type delivery system for this indication in children and adolescents. Therefore, we have learned that biodegradable PLGA-type systems are not suitable for development of sustained release BPG treatments and that although the PCL system provides favourable release behaviour, the total size of the implant may still present a hurdle for future development.