In order to maximize the efficiency of ever increasing aluminium recycling, different mutually related aspects need to be optimised. Economic efficiency of aluminium recycling is crucially dependent on the costs, arising at different stages of recycling process. Among these, transportation costs represent very important part of overall cost balance. The general model, based on the principles of reverse logistics, was developed and applied with the aim of reducing the extent of internal aluminium scrap transportation required between certain production units of an aluminium manufacturing plant. Linear optimisation model was used for calculating minimum annual transport costs and optimal way of in-plant transport for two transport models in order to determine the most efficient option. In case of in-plant aluminium recycling only collected aluminium materials are transported which means that the most important factors in such case are fixed transport costs and variable transport costs, latter depending on distance, transported quantity, energy used for transport and other operating costs. In the first transport model, the direct transport of collected aluminium scrap from each individual source to in-plant processing units was assumed. In the second transport model, one collection site was assumed where scrap is collected and then transported to in-plant processing units. The optimisation model was also applied for determining a dependence of optimal transport model on annual quantities transported internally and on the distances between sources and processing units. It was found out that the annual transported quantities and distances between sources and in-plant processing units have a significant impact on the optimal transport model. The developed optimisation model showed that environmental and economic objectives are not always in conflict. It is also shown and commented which aspects influence reverse logistics model for in-plant recycling.