Failure times of components are traditionally used to evaluate their reliability. An alternate approach is to analyze the degradation data accumulated during the component’s testing or during its normal operation. Degradation analysis is particularly useful when it is not possible to observe a significant number of failures. This is the case for metallic Integral Fast Reactor fuel pins irradiated in Experimental Breeder Reactor II, where failures have not taken place under normal operating conditions. A degradation analysis methodology is presented and applied to these pins. The time-to-failure distribution for the fuel pins is estimated based on a fixed threshold failure model. The confidence intervals of the distribution are calculated using a parametric bootstrap method.