Cyclical industrial networks are becoming highly desirable for their efficient use of resources and capital. Progress toward this ideal can be enhanced by mimicking the structure of naturally sustainable ecological food webs (FWs). The structures of cyclic industrial networks, sometimes known as eco-industrial parks (EIPs), are compared to FWs using a variety of important structural ecological parameters. This comparison uses a comprehensive data set of 144 FWs that provides a more ecologically correct understanding of how FWs are organized than previous efforts. In conjunction, an expanded data set of 48 EIPs gives new insights into similarities and differences between the two network types. The new information shows that, at best, current EIPs are most similar to those FWs that lack the components that create a biologically desirable cyclical structure. We propose that FWs collected from 1993 onward should be used in comparisons with EIPs, given that these networks are much more likely to include important network functions that directly affect the structure. We also propose that the metrics used in an ecological analysis of EIPs be calculated from an FW matrix, as opposed to a community matrix, which, to this point, has been widely used. These new insights into the design of ecologically inspired industrial networks clarify the path toward superior material and energy cycling for environmental and financial success.