Objective - To create a stochastic model to quantify the risk that shipments of cattle from regions within the United States would contain animals seropositive for bluetongue virus and to determine shipment-level accuracy of serologic testing by use of a competitive ELISA (c-ELISA). Sample Population - 19,216 shipments containing 528,918 cattle and calves. Procedure - Data were obtained on number of animals and state of origin of cattle in export shipments originating within the United States between January 1994 and March 2002. Probability distributions for size of export shipments were determined for all states within the United States, and distributions for agar gel immunodiffusion and c-ELISA accuracy (sensitivity and specificity) were determined from expert opinion and review of the literature. The model simulated selection of a shipment and then determined the probability that a threshold number or percentage of cattle within that shipment would have a positive c-ELISA result. Shipment-level sensitivity, specificity, positive-predictive value, and negative-predictive value were calculated. Results - Substantial differences were evident in the regional probability of a shipment being declared positive, with shipments from northeastern states having the lowest probability and shipments from southwestern states having the highest probability. The c-ELISA had variable predictive values at the shipment level, depending on the threshold used and the prevalence of antibody-positive cattle within the region. Conclusions and Clinical Relevance - Results from this study will aid importers in making scientifically based decisions regarding risk of importing antibody-positive cattle.