FDA recommends including a control group within the animal study design, or an explanation as to why a control group was not included. Additionally, when considering the number of animals needed to generate sufficient data that can support the safety and performance of a medical device, it is important to utilize sufficient animal numbers to obtain predictive outcomes. FDA believes that this determination can best be made after bench testing is complete and the device iterations are finalized. FDA strongly recommends that you conduct definitive animal studies on the market ready device except as required to scale, if needed, to implant in the animal model. The number of animals in the study should be based on sound scientific justification with consideration for the difficulty of the model and whether one or more test articles and/or control articles can be reasonably studied in a single animal. For example, FDA believes that deployment and handling studies can often be performed multiple times in the same test subject, or incorporated into a chronic safety study. By contrast, studies involving high-risk implants such as prosthetic joints can involve a high degree of expertise and some expected morbidity, such that a relatively large number of animals may be appropriate in order to establish device safety. Based on our experience, typical animal studies in a higher species (e.g. sheep, goat, nonhuman primate) generally have 3-9 animals per group/time point. However, in all cases a scientific justification should be provided in the protocol for the numbers used. FDA encourages you to discuss proposed animal studies, including the number of animals to be involved, prior to implementation through the Pre-Submission process.