A symposium discussing the implications of certain phenomena observed in radiation biology for cancer risk assessment in general. In July of 2002 a workshop was convened that explored some of the intercellular phenomena that appear to condition responses to carcinogen exposure. Effects that result from communication between cells that appear to either increase the sphere of damage or to modify the sensitivity of cells to further damage were of particular interest. Much of the discussion focused on the effects of ionizing radiation that were transmitted from cells directly hit to cells not receiving direct exposure to radiation (bystander cells). In cell culture, increased rates of mutation, chromosomal aberration, apoptosis, genomic instability, and decreased clonogenic survival have all been observed in cells that have experienced no direct radiation. In addition, there is evidence that low doses of radiation or certain chemicals give rise to adaptive responses in which the treated cells develop resistance to the effects of high doses given in subsequent exposures. Data were presented at the workshop indicating that low dose exposure of animals to radiation and some chemicals frequently reduces the spontaneous rate of mutation in vitro and tumor responses in vivo. Finally, it was concluded that considerable improvement in understanding of how genetic variation may modify the impact of these phenomena is necessary before the risk implications can be fully appreciated. The workshop participants discussed the substantive challenge that these data present with respect to simple linear methodologies that are currently used in cancer risk assessment and attempted to identify broad strategies by which these phenomena may start to be used to refine cancer risk assessment methods in the future.