There is much talk within the simulation and training community of capturing and analysing training data to support a more quantitative approach to training management. However, how should such data be managed and protected?