This paper discusses how health risk assessments have been conducted over the past 10 years and some of their shortcomings. Improved methods and their impact on how exposure limits for air contaminants should be established are also discussed. Examples are given of possible pitfalls in all areas of the assessment process, i.e. hazard identification, dose-response assessment, exposure assessment and risk characterisation. It is concluded that, in future, more enlightened approaches to interpreting animal bioassay data should improve hazard identification, and biologically based disposition and cancer models will provide better estimates of actual cancer risk. Reliance on worst-case exposure scenarios should no longer be necessary in the light of better information on specific exposure parameters and improved data handling methods.

Support: Digital