To develop a secure single-cell RNA sequencing (scRNA-seq) classification method, which can enhance data analysis precision and ensure the security of sensitive information, and to promote the application of scRNA-seq technology in various fields. Method/Process The paper proposes a solution based on trusted execution environment(TEE). The training data is encrypted and transmitted to TEE. It is decrypted in a secure and isolated environment, while training the model to obtain the trained model parameters. Automated cell type identification using neural networks (ACTINN) and support vector machine(SVM) are used for cell classification in both TEE and traditional plaintext environments. The results are compared and analyzed. Result/Conclusion The results show that the F1 score of the two classification models in TEE environment reaches 0.904 and 0.879, respectively, which is comparable to the performance in traditional plaintext environment. The secure execution environment provided by TEE has extremely limited impact on the accuracy and efficiency of the models. This is of great significance for seeking both secure and efficient data processing solutions in scenarios where sensitive or private data needs to be processed.