The aim of this study was to investigate whether secondary science teachers had different styles of technology adoption for technology-based assessments (TBAs), and whether teachers with different adoption styles showed different patterns of beliefs, attitudes and intentions regarding TBAs. By combining the variable-centered and person-centered approaches to explore science teachers' adoption styles and associated beliefs, this study sheds light on how to provide needed supports and resources for teachers who adopt TBAs differently. We first employed latent class analysis to identify science teachers' TBA adoption styles and then utilized multigroup path analysis to examine the beliefs, attitudes, and intentions within each style. A total of 494 secondary school science teachers from northern Taiwan participated in this study. A validated questionnaire about teachers' TBA use and related technology beliefs was used to collect data. The latent class analysis revealed three different adoption styles: Technology Rich Adopter, Technology Selective Adopter, and Technology Minimalist Adopter. The path analyses of the three groups showed that while Technology Selective and Technology Minimalist Adopters' intentions to use TBAs could be predicted by their attitudes, perceived behavioral control, and subjective norms, Technology Rich Adopters' intentions were not associated with subjective norms. Additionally, whereas the same set of significant attitudinal antecedents were found among the three groups, the effects of antecedent variables on perceived behavioral control and subjective norms varied.