Learning curves have been frequently applied in production/operations management and in various logistics processes in many manufacturing and service organizations. However, studies on their integral use in the supply chain are recent. This paper is a contribution to fill this knowledge gap by measuring the impact of learning on lead time in logistics management systems. To demonstrate this, the empirical study was used as a methodological tool. The logarithmic-linear models with their terminology and calculation expressions were applied to three case studies representatives of the logistics systems proposed by the Supply Chain Operations Reference (SCOR) model: make-to-order, make-to-stock and engineer-to-order. As a result, the first two were adjusted to the Stanford model and the third to the De Jong’s model. Their learning curve, mathematical expressions and a sensitivity analysis were determined. This demonstrated the relevance of the approach used and its difference with respect to previous publications that mostly analyze links or parts of the supply chain.