The stated aims of the ICRP for this set of recommendations are to consolidate and simplify the existing system of protection and to remain generally consistent with current international and national practices. While these aims appear to have been met in the parts of the recommendations that concern biological effects and the quantities used in radiological protection, there are changes in those recommendations relating to implementation of the system of protection that need to be re-considered; specifically: (a) the consolidation of dose constraints against a benchmark of natural radiation, (b) the treatment of optimization, (c) the definition of the scope of the system of protection, including the combination of exclusion and exemption, and (d) the conservatism in environmental protection. These are considered in the following sections. (a) The consolidation of dose constraints against a benchmark of natural radiation In the draft recommendations, the dose limits are retained (with the same values) for use in protecting individuals from “all regulated sources in normal situations” but the dose constraints that are applied to individual sources will be much more important. They will be quantified for all “controllable situations” and will represent the “fundamental level of protection . . . where action to avert exposures and reduce doses is virtually certain to be justified” [para 133]. The emphasis on dose constraints seems appropriate and the clarification of the distinction between individual-related limits and source-related limits (the dose constraints) is helpful. In its most recent general recommendations, adopted in 1990, the Commission rationalized the recommended public limit on the basis of two considerations; at the recommended limit the impact on health would be very small, and the value chosen was similar to the annual dose from natural background. In the recommendations presented in this draft, natural background becomes “a benchmark for judgement about . . . relative importance [of added exposures] and the need for action” [para. 157] and the constraints are arranged in groups on a scale of multiples of annual natural background. Such a scheme may serve to emphasize the idea that different constraints are needed for different circumstances, and a consolidation of all the constraints recommended over the last decade is certainly helpful. However, the presentation relative to annual natural background (rather than in terms of risk) creates more issues than it resolves. For example, the constraints are not explicitly related to risk but the implication of the text is that potential exposures can be included in this scheme. The benchmark is in terms of annual dose but what is the logic of including constraints for one-off doses (sheltering, say) in this scheme? Furthermore, the basis for the control in the current system is that of lifetime risk. Here, the basis seems to have become an annual dose. This represents a fundamental change in approach that is neither acknowledged nor justified in the document. Although benchmarking in this way might help to de-emphasize the threat of small increments to the natural background, it tends to exaggerate the importance of incremental doses above a few mSv. It is a weak basis for the fundamental control levels; the impact of presenting the constraints in this way is to diminish the science base and reduce the credibility of the recommendations. A better approach would be a straightforward grouped tabulation of the extant dose constraints, with some rationale for the relative magnitudes. Constraints for one-off doses (for countermeasure action, for example) should be distinguished from constraints for continuing sources. Other than noting that the doses being considered are incremental to natural background there is no need to mention natural background here. (b) The treatment of optimization. One of the problems in the implementation of the current system of protection is that it is often difficult to arrive at a generally-accepted level of protection in the public area. The judgement of what is reasonably achievable in applying the ALARA principle is difficult and we have many instances where rudimentary risk analysis indicates that resources are not being well-spent from a societal view-point. The Commission is now proposing that protection should focus more on individuals and that societal (or utilitarian) aspects of protection should be of lesser importance. The intent seems to be to reduce the role of collective dose, which has proved to be a controversial tool in protection. However, the downplaying of the role of societal or collective considerations in protection is not supported by any argument, let alone a persuasive one. The process of optimization of protection is described with many different words for different situations but the underlying principle seems to be lost. Optimal protection is said to result when exposure to individuals “represents the best choice in the prevailing circumstances” [194]. The role of optimization is to “foster a ‘safety culture’” as defined in the BSS [195]. Workplace exposures are still to be “as low as reasonably achievable, social and economic factors taken into account” but the control of emission should involve the “best available technology not entailing excessive costs . . . with due consideration to social and economic factors” [194]. Since the adjectives such as “best” and “excessive” are not defined, the distinction between these expressions is illusory. In steering away from the application of collective dose, the Commission introduces the idea of the dose matrix; one assesses the number of exposed individuals, magnitude of individual doses, dose distribution in time and a variety of other characteristics of the situation. It is an extension of the previous idea that collective dose should be “disaggregated” and different weighting given to the different components. However, the principle that would guide this application has never been enunciated and, unfortunately, there is no guidance in this draft for applying the dose matrix. (What, for example, constitutes “exposed”? Given the linear model for the dose-effect relationship, what is the basis for the valuation of numbers of people and the magnitude of their doses? ) We seem to be left with very much the same kind of ALARA assessment that we do already, with the same problems, cloaked in different words. A better approach would be to lay out clearly the principle that was being applied for optimization, complementing the control achieved through the application of dose constraints. This would be the key recommendation. The issues that arise and recommendations about how such issues might be resolved could follow but expressions and procedures peculiar to particular regulatory regimes should be avoided. Such an approach should be structured within the general framework found in risk assessment methodologies around the world in which the risks are analyzed in context, options are examined for controlling the risks, and decisions to authorize made on the basis of some kind of multi-factor analysis. It would be a pity to lose this opportunity to improve the harmonization of radiological protection with protection in other areas. (c) The definition of the scope of the system of protection including the combination of exclusion and exemption The Commission attempts unsuccessfully to clarify what sources of exposure should be included in the system of protection. The system is said to apply to “controllable sources”; other sources are excluded [para. 15]. Controllable is where “either the source of exposure or the pathways leading to the doses received by individuals can be controlled by some reasonable means” [para. 15] This definition is by no means clear – what is reasonable? There is a further difficulty since there is a suggestion to exclude materials containing artificial radionuclides at concentrations below the low-end of the range of currently internationally-exempted concentrations and to exclude natural radionuclides at the high end of the natural range. This combination of exclusion and some current exemption practice may preclude or at least discourage the implementation of higher but reasonable exemption values for particular purposes – for example in NORM guidelines. A simple screening process for all sources and exposures, without any a priori exclusion and without reliance on vague definitions, would seem to be a better approach. (d) The conservatism in environmental protection Criticism that the previous position “protect man – thereby protect the environment” has prompted the Commission to adopt principles developed by the IAEA relating to sustainability, biodiversity and habitat protection and to outline an approach for protecting biota from radiation. Accordingly, the Commission defines “radiation levels where action is needed and with explicit reference to natural background rates” [para. B22]. Here, there is an implicit suggestion that there should be concern at doses to biota a few times natural background and this is made explicit in Figure B1. Such a suggestion is far too conservative, given the current understanding of the expected no effects values for biota. Further, the Commission’s aim is for consistency with human protection and the system to be developed would include, for example, the development of reference biota. These should help in deriving ambient activity levels. However, rather than focussing on developing a system that parallels that for humans and is peculiar to radiological protection, an approach that has more in common with current environmental risk assessment and management practice would be more helpful and more likely to be accepted.