Our estuaries, and the benefits that we derive from them, are threatened by the cumulative effects of interacting stressors. Separating the impacts of anthropogenic stressors from natural variability in the marine environment is extremely difficult. This is particularly true for estuaries, due to their inherent complexity and the prevalence of difficult-to-manage diffuse stressors. Successful management and protection of these valuable ecosystems requires innovative monitoring approaches that can reliably detect anthropogenic stressor impacts. In this thesis, I examined approaches for detecting the effects of three diffuse land-derived stressors (sedimentation, nutrient loading, and heavy metal contamination) on estuarine benthic communities. Using Gradient Forest analysis, I explored the relative importance of environmental factors, operating across multiple spatio-temporal scales, in influencing patterns of compositional turnover in estuarine benthic macroinvertebrate communities across New Zealand. Both land-derived stressors (sediment mud content and total sediment nitrogen and phosphorus content) and natural environmental variables (sea surface temperature, Southern Oscillation Index, and wind-wave exposure) were important predictors of compositional turnover, reflecting a matrix of processes interacting across space and time. Generalized linear models were used to link these turnover values to measures of benthic macroinvertebrate diversity, which are commonly used as indicators of ecological health. Based on compositional turnover, I could disentangle the negative effects of land-derived stressors from natural environmental variability. Critical stressor levels associated with high rates of compositional turnover were identified, providing a useful contribution to the current knowledge on land-derived stressor effects. Once I had determined that anthropogenic impacts could be disentangled from natural variability, I developed indicators (Benthic Health Models; BHMs) to assess estuary health in response to two dominant coastal stressors (sedimentation and heavy metal contamination). Benthic macroinvertebrate community data were used in separate canonical analyses of principal coordinates to create multivariate models of community responses to these stressors. Both models performed well (R² = 0.81, 0.71), and were unaffected by regional and estuarine typology differences. They offer a sensitive and standardised approach to assessing estuarine health that allows separation of the two stressors. I also examined the potential for emerging molecular approaches to inform estuary health assessment. Recent advances in environmental genomics allow characterization of less visible forms of benthic biodiversity, offering a more holistic view of the ecosystem and potentially providing early warning signals of disturbance. A manipulative nutrient enrichment experiment was conducted in two estuaries and environmental DNA (eDNA) metabarcoding was used to examine the response of eukaryotic (18S rRNA), diatom only (rbcL) and bacterial (16S rRNA) communities. Multivariate analyses demonstrated differential changes in examined communities between sites, suggesting a context dependent response to nutrient enrichment. These patterns aligned with changes in morphologically identified macroinvertebrate communities, confirming concordance between eDNA-based and current monitoring approaches. This work represents a first step towards the development of molecular estuary monitoring tools, which could transform current approaches to ecosystem health assessment. This thesis demonstrates that the detection of anthropogenic impacts on estuarine benthic communities requires an understanding of the response of communities to stressors and how this response is modified by natural environmental processes operating at different spatio-temporal scales. My research contributes to the management and protection of estuaries by improving knowledge on the processes generating broad scale patterns in benthic macroinvertebrate communities, developing indicators that can be used to assess estuary health and demonstrating the potential of eDNA metabarcoding as a new tool for estuary health assessment.