This comment refers specifically to section 2.5, on effective dose. In my view, ambiguities remain regarding the extent to which a dose assessment can make use of non-default model parameter values while remaining within the definition of effective dose. The recommended approach should therefore be clarified. In paragraph 4, it is stated that: “In the assessment of effective dose … changes may … be made to the assumed physical and chemical characteristics of inhaled or ingested radionuclides …”, and also that “… specific biokinetic data may also be used …”. Varying such parameter values not only affects the intake that would be determined from a set of bioassay measurements, but also the doses (absorbed, equivalent and effective) determined from that intake. It follows that paragraph 4 is in apparent contradiction to the statement in paragraph 11: “Because effective dose is calculated using reference parameter values and applies to a reference person, no uncertainty is attached to the reference dose coefficients”. Paragraph 11 implies that no parameter values that relate intake to dose can be varied, whereas the parameters of the general type identified in paragraph 4 as being potentially variable would also affect the dose coefficient. Furthermore, if parameter values can be varied within the calculations of intake from measurement and dose from intake, then the uncertainties associated with these parameter values will propagate through to both intake and dose. It has been suggested that a parameter value could be allowed to be varied in the calculation of intake from measurement, but fixed at its default value in the calculation of dose from intake. However, it would be logically inconsistent to vary a parameter value in one stage of a calculation and then fix it in the next stage. Also, in some circumstances, this practice would result in overestimates of dose. A better solution would be to make it clear that parameter values describing physical and chemical characteristics of the material can be varied in both stages of the dose calculation, whereas parameter values describing the characteristics of the reference person for whom the dose is calculated would always be set at default values. Material-specific parameters would include parameters describing the particle size distribution of an inhaled aerosol, and absorption Types or absorption parameters describing uptake from the respiratory tract to body fluids. Parameters specific to the reference person would include those describing particle transport in the respiratory tract, and organ and body masses. Some parameters could be considered to be both material-specific and reference person-specific. Examples include systemic model parameters; these are probably best considered to be reference person-specific since the main source of uncertainty is likely to be variability between individuals, one instance being the long term retention half-time of caesium. Given that there is some scope for ambiguity, and since dosimetry services need clarity on this issue when defining procedures for the assessment of effective dose, it would be appropriate to tabulate those material-specific parameters that can be varied from their default values in an assessment of intake and dose.