During the last years, a series of studies on combinations of oestrogenic, thyroid-disrupting and anti-androgenic chemicals at low doses have been published. The available experimental evidence shows that combination effects may result from endocrine disrupters that each produces very small effects, if they are present in sufficiently large numbers. This review examines the implications of these findings for chemicals risk assessment and epidemiology. It is concluded that a lack of knowledge about relevant exposure scenarios presents serious obstacles for better human risk assessment. Epidemiology needs to abandon its focus on single endocrine disrupters and has to embrace the reality of endocrine disrupter mixture effects by developing biomarkers that capture cumulative exposure to endocrine disrupters.