I would like to thank the Commission for the opportunity to comment on the draft 2005 recommendations. The opinions which follow are mine and do not reflect those of any organization or government agency. 1. The system as proposed could be improved to better address Commission recommendations in (S3) for simplicity and wide applicability. The system should consist of a single limit based on avoiding acute health effects and a process for optimization to arrive at constraints based on reducing potential long term effects. The current draft could be easily modified to this purpose by identifying the highest dose value in Table S1 as the single dose limit, and the other values as recommended constraints for specific exposure situations. 2. The system should make it clear that two general principles apply: (1) limitation of dose based on avoidance of deterministic effects and acute injury and (2) further reduction, through optimization, to reduce potential stochastic effects. The first principle implies a limit for individuals, rather than sources, and applying to all exposures. It should be identified as a limit (rather than a constraint) since it represents avoidance of predictable health effects, provides “a level of protection for individuals that should be considered obligatory”, and should apply to all circumstances. The limit should be expressed in dose, but based on risk of acute or deterministic effects. Consideration should be given to expressing the dose limit both in terms of an acute dose limit, and a corresponding value for chronic exposure. 3. The second principle (reduction of potential stochastic effects) is the basis for optimization and recognizes the fact that potential stochastic effects are assumed but not known over much of the lower range of exposure. “Constraints” should be identified as those levels arrived at through optimization. The current draft’s use of multiple levels for limitation of maximum dose confuses the distinction between limitations based on deterministic effects, and lower levels optimized to limit potential stochastic effects. There should be a single limit. Lower values identified (e.g.20 mSv/y in Table S1) are necessarily constraints. In Table S1, they are constraints based on exposure situations. The Commission may want to identify constraints based on exposure situations as appropriate for specific dose-based recommendations from the Commission. Note that emphasizing the situation-specific nature of constraints removes the distinction between interventions and practices. Since interventions and practices take place in different circumstances, it should be expected that different constraints could apply. 4. Constraints should be based on dose only, since the risk in that range is much less certain. The point of constraints is to benchmark exposure reduction. That reduction is best measured against the metric of dose, which has some real physical and biological meaning, as opposed to risk which is a largely modeled and extrapolated construct. 5. Source-specific constraints should not be specified since the particulars of the optimization process are too varied to be knowable in the general context of these recommendations. As a simplifying tool, a dose level could be proposed as representing “presumptively optimized” levels. Rather than being an upper level on constraints, this would be a level low enough to be presumed protective without further analysis (e.g. 0.01 mSv/y). 6. The system should be broad and flexible enough to include all sources of exposure including occupational exposure and exposure to technologically-enhanced sources. Exclusions should only be made for specific exposure situations, not for sources or concentrations. The only exposure situations that should be excluded are medical exposures to the patient, exposures from naturally-occurring sources not influenced by human activities, and emergency lifesaving efforts. On this basis, the exclusion levels in Table S2 are unnecessary.