Central to the development of rational trade policies pertaining to bluetongue virus (BTV) infection is determination of the risk posed by ruminants previously exposed to the virus. Precise determination of the maximal duration of infectious viremia is essential to the development of an appropriate quarantine period prior to movement of animals from BTV-endemic to BTV-free regions. The objective of this study was to predict the duration of detectable viremia in BTV-infected cattle using a probabilistic modeling analysis of existing data. Data on the duration of detectable viremia in cattle were obtained from previously published studies. Data sets were created from a large field study of naturally infected cattle in Australia and from experimental infections of cattle with Australian and US serotypes of BTV. Probability distributions were fitted to the pooled empirical data, and the 3 probability distributions that provided the best fit to the data were the gamma, Weibull, and lognormal probability distributions. These asymmetric probability distributions are often well suited for decay processes, such as the time to termination of detectable viremia. The analyses indicated a > 99% probability of detectable BTV viremia ceasing after ≤ 9 weeks of infection in adult cattle and after a slightly longer interval in BTV-infected, colostrum-deprived newborn calves.