Seismic base isolation is widely used to ensure desirable performance of bridge structures against earthquakes. Seismic response of base-isolated bridges is strongly dependent on seismic isolator’s properties. However, these properties vary under the effect of multiple conditions such as temperature. The bounding analysis approach, recommended by current design codes to consider such variations, does not rigorously take into account the probability of a simultaneous occurrence of earthquakes with these conditions. As a result, base-isolated bridge design may not be optimised and the reliability of base-isolated bridges remains uncontrolled. This paper presents a probability-based reliability assessment method to consider the variations of seismic isolation properties under different conditions. Temperature, seismic hazard, the dimensions and the material mechanical properties of key structure elements are modelled as random variables. An application of the methodology is demonstrated through a case study of a base-isolated two span reinforced concrete bridge. Two limit states are considered: (1) in terms of the bending moment capacity of the bridge’s pile and (2) in terms of the displacement capacity of the seismic isolator. Preliminary results reveal that, for the case-study bridge, the global reliability of the bridge is equal to the reliability of the limit state (2). This is because the reserve provided by the bias and security factors is applied in the case of the limit state (1) but not considered for the limit state (2). What’s more, results reveal that for the case-study bridge, the seismic reliability depends on the seismic input severity more than on low temperature severity. This paper advances the basics of a method to assess the reliability of base-isolated bridges. Its application is demonstrated and preliminary results are obtained through a case study. A more systematic study using the proposed methodology should allow for the establishment of seismic guidelines with uniform target reliability.