Surgical training programs are evolving towards competency-based training, which relies on an increased frequency and type of assessment to make informed training, progression and certification decisions. Any assessment system is vulnerable to bias from many sources, so it is important that these biases be explored, documented and minimised wherever possible. This presentation focuses on the sources of assessment bias in the Australian surgical training environment, using examples from data available from the specialty training organisations associated with RACS, including the AOA 21 orthopaedic training evaluation. Suggestions for mitigating and reducing these biases will also be discussed.