The data models for the structural experiments organize formally the information involved in the structural experiments, and the formally organized information is used for implementing the computer systems. Previous research on the data models has provided various data models for their own application fields of the structural experiments, but their development methods has not been carefully reviewed and discussed. This paper analyzes the development and implementation procedure of the data models for structural experiments including the Lehigh Model for large-scale and hybrid structural experiments, the NEES data model for general categories of experiments, and the KOCED data model for distributed and shared experiments. Based on the analysis, some noticeable suggestions are given to be considered for developing the comprehensive data models for the structural experiments. The domain of the data models should be clearly defined at the initial stage of the development procedure. The categories and the levels of the information need to be decided considering the characteristics of the structural experiments. The most significant part of a data model is to be formulated first because the initial formulation affects the rest of the development procedure and the data model evolves during the development procedure.