In order to alleviate the problems of low efficiency of online trading matching of aquatic products, high security vulnerabilities of private information leakage and commodity information fraud between trading parties, and high supervision cost of online trading platform, a trading matching model for aquatic products based on blockchain technology was proposed, considering the characteristics of multi-attribute matching in the trading process of aquatic products. In this model, the key information in the trading process of aquatic products was transferred into the blockchain, and the decentralization, credibility and non-tamperability of aquatic products trading were realized. In order to maximize the matching accuracy and efficiency, an improved ant colony optimization algorithm was employed to derive the global optimal matching solution. Then in order to test the effectiveness of the trading matching model, a prototype system was designed and implemented by taking Hyperledger Fabric as the underlying architecture. The research results showed that the trading matching model based on blockchain can meet the needs of both aquatic products trading parties and trading platforms in terms of cost, efficiency and security. When the cumulative number of supply and demand orders submitted to the system within two hours was less than 1626, it did not exceed the upper limit of matching time specified in the model, and the trading matching system can operate normally and smoothly. This amount of data can basically meet the demand of online trading volume of B2B platform.