Software testing is a crucial means to ensure the quality and reliability of software, playing an indispensable role in the software development process. The design of test cases is a pivotal process within software testing. Well-designed test cases are a prerequisite for effective software testing, as the quality of test code determines the overall testing quality and effectiveness. With continuous software version iterations, a rich history of test cases accumulates during the testing process. As the functionality of the system grows, the number of required test cases and lines of code also increases, leading to reduced efficiency and longer testing cycles for testers. Test case reuse involves utilizing accumulated historical test cases and related materials for new testing tasks. Leveraging historical test cases effectively to enhance new testing tasks and improve tester efficiency is a significant challenge. In response to the aforementioned issues, this paper proposes and implements a test case reuse method based on deep semantic matching. The method employs natural language processing techniques to extract textual information and vectorize key fields of test cases, establishing a quick retrieval index. Upon receiving a user input describing the testing functionality, the method utilizes deep semantic matching technology to recommend suitable historical test cases. We improve the test case reuse method based on a deep semantic matching model and adopt the negative sampling method to increase the robustness of the model. The method's effectiveness is tested using three datasets from the text matching domain, with the Spearman coefficient as the evaluation metric, demonstrating superior matching performance compared to other baseline algorithms.