Comments on 2005 Draft ICRP Recommendations. The previous recommendations of the ICRP were that the three fundamental principles of radiation protection should be Justification, Optimization and Limitation. These principles have gained widespread acceptance among operators of nuclear installations, users of radioactive substances and other sources of ionizing radiation, and regulatory authorities. While it is true there have been cases of misunderstanding and misuse of these principles, these have been due largely to a failure of the former recommendations adequately to explain the rationale and application of the principles. The principles are now widely understood and any attempt substantially to modify them should be made only if a demonstrable improvement in radiation protection can be certain to result. The 2005 Draft Recommendations do not completely remove the principle of justification but it has been substantially de-emphasized; this reduction in emphasis will make it more difficult for regulatory authorities to gain full co-operation from intending users of sources of ionizing radiation. The proposed removal of collective dose as a recommended tool for successful optimization represents another retrograde step. The replacement of collective dose by the conceptually much more difficult "dose matrix" does not seem likely to lead to better optimization procedures and will not assist regulatory authorities in improving radiation protection. Collective dose had caused difficulties in application, particularly in the period immediately after its introduction in ICRP Recommendations. These problems appear to have become less in recent years. I believe that a clearer explanation of the use of the concept and possibly the inclusion of a defined period of some centuries after which collective dose calculations should be truncated would have been sufficient to eliminate remaining difficulties, apart from wilful misuse. The replacement of dose limits by dose constraints as one of the fundamental principles does nothing to improve radiation protection. Dose constraints have been widely used as a planning tool and have proved valuable in that capacity. A fuller explanation and clarification of their use in the revised Recommendations would have been valuable; the replacement of limits by constraints will inevitably lead to confusion even among professionals in the field of radiation protection. Exclusion at any level is not appropriate for artificial radionuclides. Exemption on the basis of trivial dose ( less than 10 microsievert per year ) would be a more valid basis for setting maximum levels for exemption. The proposed exclusion levels could give rise to considerable misunderstandings and misuse, particularly if they were to be rigourously applied, for example in the case of building materials where there could be a potential conflict with existing European recommendations. Effective dose is a useful concept and should not be replaced or redefined. It could with advantage be better explained, especially as regards its use in inappropriate situations. The changes in radiation weighting factors, tissue weighting factors and risk coefficients all appear to reflect improvements in scientific knowledge and, as such, are to be welcomed. However, it is not helpful to make these changes in advance of the publication of the supporting data. This is particularly so with regard to the epidemiological evidence. The specific inclusion of the environment in the draft Recommendations is welcome. It is probably true that protection of man is sufficient to ensure protection of the environment, but this requires confirmation in particular circumstances, especially as regards waste disposal. However, the rather elaborate procedure set out in the draft Recommendations appears un-necessarily complex. The selection of representative plants and animals for detailed consideration is premature and should be preceded by a more general consideration of the environmental impact of sources of ionizing radiation.