Scientific data are strategic resources, and the aggregation of scientific data is an important method to seize the upstream and competitive highlands of scientific data. Notably, it is challenging to grasp the international situation and the scientific laws concerning the mode of scientific data aggregation; exploring the modes and methods of scientific data aggregation that are suitable for China's national conditions is also difficult. This paper investigated and analyzed the modes of scientific data aggregation both at home and abroad from the viewpoints of international organizations, international scientific programs, government agencies, and professional data centers. Five modes of scientific data aggregation were summarized, including scientific research projects converging to designated data centers/repositories, scientific research projects dispersing to data centers/repositories, individual scientists submitting datasets to data centers/repositories with published papers, scientific research projects/individual scientists sharing directories/networks, big data computing/processing platform, and citizen science models of open and public convergence. This paper analyzed each mode and the corresponding cases. On this basis, the paper put forward six suggestions for the reasonable aggregation of scientific data in China, including the implementation of the “Measurement of Scientific Data Management”, certification of data aggregation centers, scientific data collection and publishing in journals, construction of data aggregation networks, aggregation of international resources, and construction of the whole data aggregation chain.