With the rapid development of big data technology and applications, sharing and supervision of massive data have become the demand of industry development. As an emerging technology, blockchain, with excellent characteristics, can promote the prosperity and development of the dataāsharing industry. However, the existing blockchain may lead to some privacy and security issues due to its transparency and lack of supervision. Hence, a secure data sharing model with supervision and privacy protection, named DBSDS, was proposed, which supports illegal content supervision and users revocation, and possesses flexible and fineāgrained access control on data sharing. Firstly, it introduces a dualāblockchain architecture, one blockchain is used to ensure the integrity and traceability of private data and another is utilized to supervise illegal users so that they can no longer upload and share private data. Secondly, a key tracing tree was built and correspondingly different key generation strategies for different users were designed, which endows regulators with supervision capabilities. Finally, combined ABPERāKS algorithm, both privacy protection and fineāgrained access control on private data can be accomplished. Furthermore, security analysis and performance comparison manifest that DBSDS is safe and owns better performance, and scheme implementation and experimental analysis indicate the practicality of DBSDS.