Sustainability has three dimensions, including social, economic, and environmental dimensions. However, existing road network design studies only focus on one or at most two dimensions, which do not allow decision makers to consider social, economic, and environmental impacts on human simultaneously. This paper proposes a multi-objective bilevel optimization model to consider all three dimensions in road network design. To examine the effect of road network design on landowner inequity and intergeneration inequity, land-use transportation interaction over time is also captured in the model. The variance of discounted landowner profit and the variance of discounted generalized user cost over time are proposed as sustainability indicators of landowner inequity and intergeneration inequity respectively. Artificial bee colony algorithm (ABC) is proposed to search the network design solutions of the upper level problem, while the method of successive averages (MSA) and the Frank-Wolfe algorithm are adopted to solve the lower-level time-dependent land-use transportation problem. Numerical studies are set up to illustrate the tradeoff between the three dimensions of sustainability objectives, the performance of the proposed algorithm, and the existence of landowner inequity and spatial inequity of residents.