General As someone who has on occasion been critical of the direction that ICRP appeared to be taking during the process of revision of the basic recommendations, I am very pleased to be able to express my complements to ICRP on the latest draft for comment. This sets out its objectives very clearly in the Preface and, in my view; the draft achieves them. Scope The separation into planned, emergency and existing situations is useful but has not quite been carried through fully and consistently. In (163) it states that these replace the previous two categories 'practices' and interventions'. However there are several places where the old categories are retained. For example in (188) planned situations are explained as being practices. Either the replacement of the old terms should be consistent or else, when the new terms are introduced, an explanation should be give as to why and under what circumstances, the old terms should continue to be used. This comment does not apply to the medical use of 'practice', which would be unequivocal if the old ICRP use was dropped. In the early Chapters emergencies and existing situations are dealt with clearly and well separated. This is not the case for Chapter 9 which is not so clear and well written as the early part of the draft. It purports to deal with both (although separation might be better) but Sections 9.3, 9.4 and 9.5 deal almost exclusively with emergencies. Chapter 9 needs serious attention. Exclusion and exemption Exclusion appears to be clear. However exemption is still a bit confusing. It should be clear that exemption is a regulatory decision that is the optimum approach. The decision may be assisted by the criteria in (46). In these criteria, the third should be separated into two as the need for justification has nothing in common with the need for inherent safety of the source. Justification and Optimisation I am pleased that justification is retained and I believe the treatment is appropriate in all contexts. Optimisation is treated in an up-to-date fashion e.g.(194) but the revised statement of principle (185) does not seem to improve on the original as restated in (193). The important concept in (227) could possibly be moved into the general section 5.8. The treatment of collective dose seems to meet earlier objections. Dose limits The retention of occupational and public dose limits is welcome, but it seems a shame that the opportunity to simplify both to annual limits has not been taken. Constraints Thu use of constraints in various circumstances has been greatly clarified. However there remains the likelihood of their being regarded as limits and unchangeable, it would help if the need to keep constraints under review - especially those set at the design stage - could be emphasised. Natural sources. This approach to fully integrate protection against natural sources is good. However the material in section 7.1 in particular is tutorial and not appropriate for this document. Paras (296/7) could be moved to Chapter 3. Protection of the environment This level of principle is appropriate for this document and at this stage in the development of concepts. Implementation of recommendations The introduction of Chapter 11 is greatly to be welcomed. It brings together a number of important practical ideas that hitherto had to be sought for throughout the recommendations. Drafting points. (171) The intent of 'other workers are treated as members of the general public' is not clear. Does it mean their dose limit is 1 mSv? (248) 'This procedure' presumably means justification although it is not clear with the references to procedures in (247), however the para seems to relate to optimisation. (269) The para is OK but the heading 6.4.2 Medico-legal may cause confusion given the controversy over what this term means. I suggest deleting the heading. (280) line 4 'or' should be 'and'. (299) Working level moth(!) introduced without definition? (311) Somewhat odd reference to the BSS in an ICRP document. (332) This list of accidents is inappropriate. (335-350) No detailed comments on these. They are repetitive and the Chapter needs considerable work. Geoffrey Webb