This study proposes a two-stage stochastic programming model to determine an optimal set of bus stations that minimizes operational, environmental, and social costs under uncertain weather conditions and customer perceptions on sustainability. The first stage of the proposed model focuses on the derivation of a set of bus stations under uncertain demand and weather conditions. Then, the second stage determines an optimal vehicle capacity (i.e., bus size) to minimize the impact of vehicle shortages. In the proposed model, different customer perceptions on sustainability are conceptualized through a range of dissatisfaction levels. Weather conditions are considered as causing higher dissatisfaction for vehicle shortages in certain seasons. The proposed model is applied to a numerical case study for a bus transit network in a college town. This study also analyzes the effect of human behavior on system costs by comparing the proposed model with a traditional approach. The results provide managerial insights on the fact that bus transit network design problems should allow for tradeoffs between different types of costs.