[en] Experience in the nuclear industry has shown that the most FR-equent route for significant plutonium deposition in man is by inhalation. After an accidental inhalation exposure to plutonium oxide has occurred, an attempt is usually made to determine the lung burden using in-vivo measurement techniques. However, this will not be possible if the amount deposited is less than the detection capability of the counter. Likewise, it may not be possible for even larger intakes if they are first discovered as a results of a routine examination after much of the plutonium has transferred out of the lungs. Under these conditions, the initial lung burden and dose assessment may, of necessity, be based entirely on excretion date. This paper discusses three plutonium inhalation cases for which autopsy data are available to demonstrate the difficulties and uncertainties of such assessments and to emphasize the need for enhanced routine surveillance of the worker and work place to assure that intakes are detected at the time they occur. An in-depth review of the findings for these three cases suggests that the use of urine excretion data applied to ICRP lung model assumption will result in large uncertainties in estimates of the pulmonary burden and lung dose. These estimates will be improved if the material inhaled can be better characterized and by using clearance half times developed for the individual rather than the default values for Class Y material recommended by the ICRP