Abstract: The central focus is devoted to exploring the robust reliability methodology for underground structures based on the Information-Gap (I-G) theory. Firstly, according to the influences of the possible ranges and the various fluctuations of uncertain variables involved in the output model on the safety performance of underground structures, the I-G model is used to handle the uncertain variables quantitatively. Secondly, the correlation between the response value and the critical value for the output model is developed by use of the failure criterion utilized for underground structures, and the robustness function is then established. On this basis, the greatest horizon of the fluctuation of uncertain variables that the structural performance can suffer is further defined as the robust reliability index which will be consistent with no-failure to measure the structural reliability. Hence, a robust reliability methodology for underground structures by means of the I-G theory is presented. Finally, a scheme to demonstrate the robust reliability evaluation for an example of the tunnel lining is particularly implemented by the proposed approach. The illustrative practical example reveals that the current robust index for structural performance has the capacity to be actively immune to unanticipated variations of uncertainty parameters. The proposed methodology affords a novel route for the reliability assessment of underground structures when the uncertain information is unavailable or scarce.