Abstract In response to the rising public concerns on environmental issues and the recently imposed governmental legislations and directives in this regard, companies nowadays ought to constantly adopt sustainable practices in their day-to-day operations. To that end, this paper seeks to optimize the replenishment strategy for a firm (e.g. retailer), who is facing a discrete time-varying demand for a temperature-sensitive (cold) product, while taking into account carbon emissions resulting from transportation and storage activities. In essence, the problem tackled herein is a variant of the classical dynamic lot-sizing problem (DLSP) for cold products with limited shelf life. To better resemble reality, we assume a stepwise cost function that is proportional to the number of fixed-capacity refrigerated trucks and cold stores used. We present three mathematical models, with the first being a cost minimization model, the second a carbon footprint minimization model, and the third a hybrid model that utilizes the carbon tax policy to jointly optimize both economic and environmental measures. We identify several structural properties of the optimal solution to the three models leading to the development of an exact solution algorithm based on the dynamic programming approach. The results of the computational experiments suggest that accounting for environmental aspects induces a change in the cost minimizing lot-sizing policy, resulting in a minor increase in the operational cost where the savings in the CO 2 related costs outweigh this minor increase. The two-way sensitivity analysis, conducted under randomized values of the demand, truck sizes, and cold stores capacities, indicates that higher savings in cost and carbon reductions are realized for increasing values of the carbon tax rate. Furthermore, on average, more cost savings and lower carbon emissions are also achieved when the values of the shelf life, truck size, and cold store capacity are increased. [ABSTRACT FROM AUTHOR]