Following an accident, it is essential to provide personal radiation monitors to the affected population, because direct personal measurements are much more reliable than reconstructing such doses through environmental modelling and assumptions about personal behaviour.

Early and follow-up direct measurements of levels of dose and incorporation of radionuclides in members of the public are the most valuable way of obtaining information on the radiological impact of an accident and for providing public information. In particular, measurements of 131I in thyroid should be taken as soon as possible following releases to the environment, owing to its short half-life. If direct measurements are not available, modelling and assumptions of transfer parameters may be required to estimate the doses received as long as its limitations in accuracy are recognized. When estimating doses using modelling techniques that are based on environmental measurements and estimates of the amount of radionuclides released, the relatively large uncertainties associated with these estimates should be recognized and clearly explained.