NIRS is the information and networking center for citizens and environmental organizations concerned about nuclear power, radioactive waste, radiation, and sustainable energy issues. We submit the following comments on ICRP DRAFT RECOMMENDATIONS 2005. These comments follow other comments from NIRS: • Reject the notion of a threshold. We caution that, even though certain diseases may appear to have a threshold, evidence for threshold doses of radiation do not exist for all radiation induced disease and since no one can accurately predict what disease will strike which individual, any perceived threshold affords us very little when it comes to protecting those most vulnerable. • Reject the notion of hormesis. We are not sure of the final health outcome despite a perceived hormetic effect. Additionally, hormetic effects are seen only for a short duration and only in very specific dose circumstances not reflective of internal, constant, low-doses. • Encourage more use of biodosimetric measurement since the best measure of damage to an individual is an Actual measure, Not a dose reconstruction. Although this technique is more appropriate to health studies and disease assessments, the ICRP must encourage it since these studies are used by ICRP for assessment of risk • ICRP has underestimated risks of radionuclides that show unexpected properties, such as strontium-90, tritium and auger emitters. Not only is more research needed, the ICRP should increase the risk associated with these radionuclides and assume a precautionary approach to be more protective. • ICRP needs to account for both genomic instability and bystander effects. Both effects see more damage at low internal doses. • Likewise, ICRP needs to realize that the mechanisms for genomic instability, bystander effect and mini-satellite mutations may explain effects seen at low doses that were not expected and are not found at higher doses an that these effects vary by cell type and individual. • ICRP stating that there is not enough known about these effects to warrant inclusion in risk estimates is a weak argument. Historically, ICRP has underestimated risk only to have to increase damage estimates later. We urge the ICRP to learn from past mistakes and take a precautionary approach when it thinks data is not fully understood or does not present a fully comprehensible picture. • ICRP dosimetric models need to account for insufficient modeling of damage to cells and molecules particularly for internal short-range radiation. Dose level formulas used by ICRP could easily become meaningless at the cellular and sub-cellular level because even a single track becomes a large dose in this dimension. A dose such as this could start a cancer or other disease, yet register as a miniscule whole body dose. • ICRP needs to use and support data that fall outside their expected model outcome. Rather than discount observation that fails to fit any favored theories science should investigate why certain data don’t fit expectations. Especially in light of phenomena like bystander effect, ICRP discounts unexpected data at its (and OUR) peril. If more cancers are seen in a study than are expected, for instance, rather than say that radiation couldn’t be responsible, maybe we should consider that the risk estimates are wrong. We call your attention to this common occurrence and ask for a reasonable consideration of inductive versus deductive methodologies. • ICRP needs to recognize that all epidemiological studies they rely on for risk estimates are subject to inappropriate data mixing, limiting assumptions about the linear-no-threshold theory (point above) and problems with recording human disease. Epi studies are a tool, not gospel, and should be regarded as such. No one study is representative of all exposure aspects, human, animal, cell response, etc. • Regarding Table S1, Maximum Dose Constraint of 100 mSv per year Effective Dose:10 rem exposure per year produces a cancer risk over a 30-yr exposure period of 1 in 4. Such a risk level is grossly outside what can be considered acceptable. Over a 70-year lifetime of such exposure the risk to exposed individuals would exceed 50%, ie greater than half of all people exposed would get cancer from that exposure. These estimates are based on official agency risk figures (see US Federal Guidance Report 13 for radiation). Using ICRP’s own risk figures produces similar results (1 in 5). 10-rem per year over 30 years of exposure is roughly equivalent to a dose received from 50,000 chest x-rays. Considering the conclusions reached by both the CERRIE Report and the CERRIE Minority Report, research must be continued, and, in the interim, existing standards must not be weakened. Precaution demands, instead, that standards need to be all the more restrictive, and that radioactive materials and wastes already deregulated (and the unregulated ones) should be brought under control.