Purpose Building information modeling (BIM) is recognized as a major innovation in the architecture, engineering, and construction (AEC) industry. Understanding the factors that influence the AEC’s adoption of BIM will benefit the research and practice of BIM. The paper aims to discuss these issues. Design/methodology/approach This study provides empirical evidence for the accumulated knowledge of BIM adoption by examining the context of Chinese construction industry. Based on the technology-organization-environment (TOE) framework in the innovation diffusion literature, the authors develop a research model that integrates the critical success factors related to the technology of BIM, the construction company and the environment in Chinese construction industry. The authors collected two different data sets from engineering consulting firms and construction firms in China, and conducted rigorous analyses using a sophisticated statistical approach. Findings The authors found that the relative advantage of BIM was a major factor that enabled BIM adoption, while the complexity of BIM was an inhibiter. In addition, management support was also a significant antecedent of BIM adoption. However, organizational readiness was significant for engineering consulting firms but not for construction firms. Surprisingly, the authors did not find consistent significant impacts of any environmental factors. Last, younger firms were more likely to adopt BIM. Originality/value One of the first to apply the TOE framework to integrate three groups of factors that may explain BIM adoption in China. Such a comprehensive framework provides a much broader perspective of BIM adoption to evaluate the impacts of different antecedent factors. The authors conducted an empirical study based on survey data collected from two different types of companies, i.e., engineering consulting firms and construction firms, representing the two parties in the principal-agent relationship of a construction project. One of the first to apply a sophisticated statistical approach, i.e., partial least squares, to analyze the data in the BIM literature.