In this study we discuss quantitative models for the optimal design of manufacturing systems. The optimization criterion chosen is the logistic performance of the manufacturing system. Three kinds of design problems are discussed: machine allocation, traffic allocation, and speed selection. In machine allocation we consider the optimal distribution of capacity, in terms of the number of machines, over a set of possible workstations. Several algorithms to solve the relevant problems are presented. The algorithms are applied to two manufacturing systems which are taken from practice. The results from these two examples show that in general the performance of a manufacturing system can greatly be improved if a system is designed with the help of the proposed quantitative techniques. In traffic allocation the optimal distribution of product demand over the workstations is treated. For a specific manufacturing system, a so-called flexible flow line, we show that improvement in terms of logistic performance can be achieved by applying the traffic allocation algorithms presented. Finally, speed selection concerns the problem of finding the speeds of the machines to arrive at an optimal trade-off between costs and logistic performance of the system. For several manufacturing environments we present algorithms for the related speed selection problems. For these algorithms we specifically discuss the computational complexity.