Purpose - This study reviews articles in the dynamic capability (DC) literature and proposes an alternative method, key-route main path analysis, to reduce the effect of sample selection bias in prior research approaches. Design/methodology/approach -This research uses the keywords ＂dynamic capabilities＂ and ＂dynamic capability＂ to collect literature from the WOS and Scopus databases, building a citation network. By combining the main path analysis and test-mining for literature review, this study reveals the global main path and branch paths. Findings-This study discovers that the development trajectory can be separated into three phases: the initial phase, the diverged phase, and the converged phase. There are two branches in the diverged phase: major main path (Macro/Collected-orientation) and minor secondary main path (Micro/Cognition-orientation). Research limitations/implications -Web of Science (WOS) and Scopus are used for searching the literature, with a limited scope of journal and review articles in management and business fields. The co-word analysis is conducted with top 100 indexing terms. Practical implications/Social implications -This study shows that most previous literature focused much more on macro/collective research than on micro/cognition research. The discovery deserves more attention and further endeavors, especially in cases of cross-level integration. Originality/value -This study contributes to suggest the key-route main path of development trajectories of the DC literature and the same research methodology can be applied to other research fields.