This paper addresses the question of expertise in relation to technology assessments, arguing both that "lay expertise" is necessary and that "lay expertise" should not be interpreted as equivalent to "patient expertise". It presents findings from a prospective study of the social implications of genetic testing for susceptibility to occupational disease and injury. The findings support the view that technologies cannot be adequately assessed in isolation from the socio-political context in which they will be utilised. Interviews with those possessing non-traditional, "lay" expertise in seeking compensation for occupational disease and/or long-term disability identified a range of institutional practices which, unless they are addressed, will significantly increase the dangers of utilizing genetic testing. The paper concludes that adequate assessment of a technology can be impeded by a focus on the technology itself, particularly when the impact of a technology may be to exacerbate already problematic aspects of existing social and political institutions.