Benzene risk assessment indicates that exposure to a time-weighted average (TWA) of 1-5 parts per million (ppm) benzene in ambient air for 40 years is associated with an increased risk of acute myeloid leukemia. Decreased white blood cell count, platelet count and other hematological indices have also been observed in persons exposed to as low as 1 ppm airborne benzene. Evidence from studies worldwide consistently shows elevated levels of benzene biomarkers that are equivalent to 0.1-2 ppm benzene in ambient air, or even higher in the general population without occupational exposure to benzene (including children). The public health significance of these observations depends on to what extent these levels reflect actual benzene exposure, and whether such exposures are life-long or at least occur frequently enough to pose a possible health threat. We reviewed the evidence and discussed possible explanations for these observations. It was concluded that while there is reason to suspect that benzene contributes significantly to elevated levels of biomarkers in the general population, there is growing concern that this cannot be definitively ascertained without concomitant consideration of the role of other factors such as metabolic polymorphisms and sources of biomarkers other than benzene, which have been insufficiently studied to date. Such studies are urgently needed for valid assessment of this potential public health problem.