The transmission service of small and topology-complex natural gas pipeline networks can be provided using an Entry/Exit(E/E) tariff mode. The booking of gas transmission capacity under the E/E tariff mode does not consider the path of gas flow, and is charged according to injections and/or deliveries. The input/output capacity at an entry/exit in the E/E mode is defined as the maximum flow that can be booked by the shippers. The pipeline company initially determines the tentative input/ output capacities at entries/exits in the pipeline network based on experience. If there exists at least one feasible operational plan for any input/output flow combination that does not exceed the capacities at the entries/exits in the pipeline network, the initially determined input/output capacities can be published and booked by shippers. The validation of input/output capacities requires finding an operational plan for the pipeline network when the actual input/output flows are uncertain parameters. Thus, the validation problem is an uncertainty optimization problem. In this work, a two-stage robust optimization model was established to transform the validation problem with uncertain parameters into a deterministic problem of verifying the worst-case scenarios. There were several sub-models in the first stage. In each sub-model, the input and output flows at the entries and exits of a gas network were decision variables, with the corresponding uncertainty set transformed into constraints. And the objective function is to minimize the inlet pressure of each compressor or delivery station to generate the worst-case scenarios. In the second stage, the minimum inlet/delivery pressures of the compressor/delivery stations under the worst-case scenarios were verified whether they meet the lower pressure limit constraints. The input and output capacities are feasible if all the constraints are satisfied. Compared to previous capacity validation methods that based on verifying multiple scenarios, the model proposed in this study eliminated the limitation of manually selecting scenarios, ensuring that the calculation results were reliable. The results indicated that in a complex pipeline network, the gas transmission cost under the E/E mode better reflects actual costs, comparing to that under the path-based tariff mode. However, the physical gas transmission capacity of a network cannot be fully utilized under this mode in most cases, resulting in capacity wastage. Based on the input/output capacities validation model, this paper explored the feasibility of applying the E/E tariff mode in regional gas pipeline networks in China, and it could also be referred for improving the open access regime for pipeline networks.