A suite of biomarkers was developed using the crab Carcinus maenas and the mussel Mytilus edulis as test organisms. The ability of the biomarkers to differentiate amongst the major toxic components and to indicate the concentration of chemical mixtures was evaluated in the laboratory. Biomarkers were also applied in a field trial and their potential to monitor environmental water quality in a chemically contaminated estuary investigated. The results from the biomarker assays were compared with and validated against two commonly used toxicity tests (Tisbe battagliai LC-50, and Microtox®). Novel methods for recognising patterns of biomarker responses were developed and assessed. The most sensitive and reliable biomarker assays investigated were neutral red retention time in crabs and mussels and heart rate and glutathione-S-transferase activity in crabs. Effects were observed at environmentally realistic concentrations; for example lysosomal enlargement was observed in mussels exposed to a complex mixture containing chemicals at environmental quality standard concentrations. Exposure concentrations required to illicit biomarker responses were similar to toxicity test EC-50 values. The ease of interpretation and clarity of the results was enhanced when data from suites of biomarkers were pooled and analysed using multivariate statistical techniques (multidimensional scaling and cluster analysis). Multivariate analysis differentiated amongst mixtures containing solely organic chemicals, metals and metal and organic chemical mixtures. Exposure response relationships to complex mixtures were established for some of the individual biomarkers tested (crab heart rate and gill metallothionein) and also for suites of biomarkers when multivariate analysis was carried out. In the field biomarkers, in both transplanted and indigenous animals, were able to differentiate between clean and contaminated sites and indicate a pollution gradient along the Tees Estuary. This was not achieved using toxicity tests. The results were displayed clearly using multivariate analysis, enhancing the power of biomarkers as monitoring tools.