Toxicity testing of drugs, pesticides, and hazardous compounds has evolved into a battery of standardized tests conducted in a range of surrogate test organisms. The toxicity of these xenobiotics in terms of their LD50 and LC50 (Dose or concentration lethal to 50% of the test population), ED50 and EC50 (Dose or concentration producing a specified response in 50% of the test population), MATC (Maximum acceptable toxicant concentration), LOEL (Lowest observable effects level), LOEC (Lowest observable effects concentration), NOEL (No observable effects level) or NOEC (No observable effects concentration) is extrapolated to humans and wildlife. Historical failures in the risk assessment process have been largely due to over reliance on regulatory toxicology and an 'assembly line' mentality to toxicology. The importance of toxicokinetics, receptor studies and biomarkers are reviewed, firstly, with reference to toxicological incidences in drug development programmes, and secondly, with reference to improved environmental risk assessment of pesticides and other contaminants. Ecological risk assessments also require multidisciplinary skills to study the entry, distribution, and biological effect and fate of chemicals to fully characterise and understand the potential adverse implications of contamination. Optimum integration of chemical measurements and biomarker responses is a challenge that will lead to an improved understanding of adverse effects and their significance in both human and ecological risk assessment. (C) 2002 Published by Elsevier Science Ireland Ltd.