The development of new and better radioactive tracers capable of detecting and characterizing osteomyelitis is an ongoing process, mainly because available tracers lack selectivity towards osteomyelitis. An integrated part of developing new tracers is the performance of in vivo tests using appropriate animal models. The available animal models for osteomyelitis are also far from ideal. Therefore, developing improved animal osteomyelitis models is as important as developing new radioactive tracers. We recently published a review on radioactive tracers. In this review, we only present and discuss osteomyelitis models. Three ethical aspects (3R) are essential when exposing experimental animals to infections. Thus, we should perform experiments in vitro rather than in vivo (Replacement), use as few animals as possible (Reduction), and impose as little pain on the animal as possible (Refinement). The gain for humans should by far exceed the disadvantages for the individual experimental animal. To this end, the translational value of animal experiments is crucial. We therefore need a robust and well-characterized animal model to evaluate new osteomyelitis tracers to be sure that unpredicted variation in the animal model does not lead to a misinterpretation of the tracer behavior. In this review, we focus on how the development of radioactive tracers relies heavily on the selection of a reliable animal model, and we base the discussions on our own experience with a porcine model.