Construction safety is important in the domains of architecture, engineering, and construction. However, studies have rarely discussed safety issues from the perspectives of employers or employees’ ignorance of safety requirements. A novel approach with modeling and reasoning functionalities is applied for semi-automated identification of safety requirements from construction safety standards. The proposed approach is based on ontological modeling and document modeling techniques. These techniques model safety concepts and requirements in semantically-rich, human-readable and computer-interpretable format. These features further enable reasoning about the concepts and requirements needed to identify applicable safety requirements. Test cases are used to validate the approach and illustrate its advantages. Validation results are discussed, and conclusions and directions for future research are given. The proposed approach can benefit construction safety by improving participants’ awareness of safety requirements.