Various random copolyesters of 3-hydroxybutyrate and 3-hydroxyvalerate (PHBV) and 3-hydroxybutyrate and l-hydroxybutyrate P(3HB-4HB) were used in the construction of biodegradable, implantable rods for the local delivery of antibiotics (Sulperazone(R) and Duocid(R)) in chronic osteomyelitis therapy. Drug loading, type of active agent, and additional coating of the implant surface all have significant contributions to the in vitro release profile. The rate and duration of Sulperaxone(R) release from P(3HB-4HB) rods were controlled by the polymer/drug ratio (drug loading). The rate of drug dissolution was substantially higher than that of polymer degradation. Therefore, the release phenomenon was more dependent on drug dissolution rather than on polymer degradation or diffusion. Coating rods with the same type of polymer substantially reduced the initial burst effect observed with the uncoated rods, and significantly decreased the release rate so that the release kinetics became almost zero order. Antibiotic release from coated rods was sustained for over a period of 2 weeks at a constant rate, whereas uncoated rods released their contents in less than a week. Impregnation of Duocid(R) into the hydrophobic polymer matrix yielded a rod with a smoother surface topography. The release from these rods was significantly higher than for rods loaded with Sulperazone(R) and a zero order release could not be obtained with these samples.