The global effect of activation cross-section uncertainties on calculated radiological quantities is investigated for the first time using a methodology based on Monte Carlo random sampling. The method is applied to the calculation of the uncertainty in the contact dose rate from the gunite shielding of the National Ignition Facility chamber after 30 yr of pulsed irradiation. Some critical cross section contributing significantly to the overall uncertainty are identified. By a reasonable reduction of the uncertainty in those cross sections, the accuracy in the calculated total contact dose rate is greatly improved.