[en] A methodology for the assessment of the uncertainties about the reliability of nuclear reactor systems described by Markov models is developed, and the uncertainties about the probability of loss of coolable core geometry (LCG) of the Clinch River Breeder Reactor (CRBR) due to shutdown system failures, are assessed. Uncertainties are expressed by assuming the failure rates, the repair rates and all other input variables of reliability analysis as random variables, distributed according to known probability density functions (pdf). The pdf of the reliability is then calculated by the moment matching technique. Two methods have been employed for the determination of the moments of the reliability: the Monte Carlo simulation; and the Taylor-series expansion. These methods are adopted to Markovian problems and compared for accuracy and efficiency