For old reinforced concrete (RC) bridges subject to environmental attack, the data for reliability assessment sometimes cannot be recorded or collected precisely due to human errors, machine errors, or some other unexpected situations. Uncertainties due to limited number of experimental data, incomplete inspection information, as well as the intrinsic randomness of random variables affect the reliability assessment of the existing RC bridges.A reliability assessment methodology under small samples is developed considering inspection results and expert experiences in this paper. The study considers the influential parameters of structural resistance as fuzzy variables.The membership function is employed to describe fuzzy characteristic of small detection samples. The fuzzy variables are transformed to the equivalent random variables and the traditional JC method is used to compute the reliability index. The proposed methodology is illustrated and validated with the reliability assessment of a fabricated RC bridge. The sensitivity of the concrete strength, the reinforcement strength and the corrosion loss on the reliability analysis are also discussed. The results show that the fuzzy range of the corrosion loss and the reinforcement strength have significant effect on structural reliability, which indicating the assessment of the reinforcing bars corrosion should be more concerned in practice measurement of the existing bridges.