Environmental DNA (eDNA) has reshaped the way we obtain ecological data, providing an unprecedented view of biodiversity. To date, eDNA biomonitoring of aquatic ecosystems has primarily focused on community composition and univariate endpoints. However, organisms do not live in insolation, and it is the collective of the community that drives many processes and helps maintain stability. Co-occurrence analysis are increasingly being used to examine the interactions between taxa. This information can then be visualised as a network with quantifiable topological features. Here I illustrate the potential of using co-occurrence networks as a biomonitoring tool. I demonstrate some field examples, highlight some of the challenges in interpreting these findings, and present some experimental evidence from a copper-spiked mesocosm study on how co-occurrence networks respond to a stressor. While our findings do add credence to the use of co-occurrence networks as a tool for assessing the responses of communities to disturbances, they also reveal the challenges associated with ecological interpretation, particularly with regards to spatial scale and sample size.