Abstract: Federated learning provides an effective way to get rid of the dilemma of “data island” , which has attracted extensive attention in academia and industry. The lack of user participation enthusiasm and the difficulty of data supervision make sustainability of federated learning cannot be guaranteed. Data trading driven by incentive mechanism is an effective way to address this problem. Firstly, the research background and significance of data trading for federated learning have been introduced. Then, the research progresses of data trading for federated learning are described. Finally, the existing problems and future research directions of data trading for federated learning are put forward.