Risk assessments use assumptions based on differing degrees of conservatism. This complicates the understanding of the uncertainty in the final risk estimate. Uncertainties arise from each component of the risk assessment process: source terms, atmospheric transport, exposure, and dose-response. Probabilistic modeling using Monte Carlo and Latin Square sampling techniques (reference to ＇＇Gambling in Latin＇＇) allows for an improved approch to risk assessment and management.