Food webs are at the heart of understanding the biological structure and functioning of ecosystems. Exploring the complex structure of food webs is one of the central issues in ecological studies. Previous studies have developed two hypotheses, i.e. link-species scaling law and constant connectance. The link-species scaling law assumes that food webs with varying diversity have the same linkage density. The constant connectance hypothesis states that food webs have constant connectance though their diversity varies. Given the inconsistency between the two hypotheses, it is critical to explore the correlation between food web complexity and diversity. Based on the constructed food web structure of a seagrass bed in the Yellow River Estuary and a well-studied actual food web dataset "TroModels", we totally collected 48 food web cases comprising estuarine, lacustrine, marine, and riverine aquatic ecosystems. In this study, the food web complexity was represented by three metrics, i.e. the number of links, linkage density, and connectance, and the number of nodes in the food web indicated its diversity. We used the trophic scaling model to describe the relationships between food web diversity and complexity in various aquatic ecosystems. Our results showed that the food web diversity and complexity of the 48 aquatic food webs varied, in which the number of nodes ranged from 4 to 124, the number of links ranged from 3 to 1830, the linkage density ranged from 0.75 to 15.71, and the connectance ranged from 0.06 to 0.25. The food web connectance differed significantly among different aquatic ecosystems (P=0.01), but there were no significant differences in the number of nodes, the number of links, and the linkage density of food webs among different aquatic ecosystems. Both of the number of links and the linkage density increased with the increasing of the number of nodes (R²=0.92, P<0.001 and R²=0.82, P<0.001, respectively), and there were no obvious correlations among different aquatic ecosystems. In addition, the food web connectance decreased with the increasing of the number of nodes (R²=0.06-0.41, P<0.001) except that the connectance of the lacustrine food webs showed nearly constant value (about 0.20) with the increasing of number of nodes. The quantitative analysis of aquatic food webs on a global scale in terms of food web diversity and complexity provides support for scientific understanding of the complex structure of food webs and the exploration of the relationship between diversity and complexity at the system scale.