The COVID-19 pandemic forced upon the world, severe social distancing restrictions, which led to prolonged confinement across populations. The latter directly impacted actors along the supply chain in a variety of industrial sectors (for instance, raw material suppliers, manufacturers, distributors, and customers, among others). Some actors involved had to cease participation altogether due to closures. As a result, the supply chain requires restructuring and its reactivation requires careful consideration. In addition to the pandemic, poor air quality has brought about an environmental crisis in recent years. Primary polluters include greenhouse gas (GHG) emissions caused by manufacturers and distributors. Therefore, this research studies the problem of restructuring a particular multicommodity and hierarchized supply chain. Specifically for companies dealing with situations derived from a reduction in manufacturing capacity and service level in light of the pandemic. In this case, a company (leader) is faced with selecting customers that it will service in pursuit of maximizing profit, all while looking to minimize GHG emissions. The consolidated demand is nearshored once the leader company decides on the customers to be supplied. That is, an order is placed on a company with a lower hierarchy (follower). The follower, in turn, aims to minimize its own manufacturing costs without exceeding the pollution limits imposed by the government. However, its manufacturing plan inevitably pollutes and incurs different costs. In addition, the follower’s decisions impact both leader’s objective functions. We propose a bi-objective bi-level programming model to study this situation. To solve the problem in reasonable computational time, a heuristic algorithm that takes into account existing asynchrony between leader and follower companies is proposed to approximate the Pareto front. Computational experimentation reveals that the proposed algorithm provides good trade-off solutions, which can reduce GHG emissions by 67% on average without significantly affecting company revenue. Moreover, the algorithm is able to provide solutions for instances of up to 1000 nodes in a competitive computational timeframe. In addition, we discuss the advantages of computing GHG emissions proposed herein. Finally, useful managerial insights are discussed by performing a sensitivity analysis regarding the distribution company’s minimum acceptable level of profit.