Airworthiness regulations are the most important reference documents for safety engineers in aviation field. For each control process, the corresponding safe constraints are included in the unstructured texts. Information extraction in an intelligent way will help safety engineers acquire specific information quickly and exactly. However, there is no such an intelligent system for airworthiness regulations until now. Recognizing named entity in the domain of airworthiness safe is an important task in constructing the intelligent system. In this paper, an entity recognition system for Chinese airworthiness regulation texts is proposed based on deep learning method. Firstly, we construct the tagging system with the consideration of multiple components in control process. The tags are classified to seven types, including organization institution, responsible person, system, document, control action, constraint condition, and regulation clauses respectively. Then, 50 airworthiness regulation texts in total are manually tagged for the seven types of named entities. Secondly, different supervised methods are adopted for the task of Named Entity Recognition (NER) to Chinese airworthiness regulation texts, including statistical method (Conditional Random Field, CRF) and deep learning method (Bidirectional Long Short-Term Memory, Bi-LSTM). The results demonstrate that the deep learning method achieves a significant performance to the task of NER for airworthiness regulation texts. It is indicated through the above analysis that the deep learning method can be utilized to construct the intelligent information extraction system, thus to promote the development of intelligent in the field of airworthiness safe.