The development of various forms of public-private partnerships for the financing, building and operating of public infrastructure has not fundamentally altered the economic calculations involved. This chapter examines to what extent it is necessary, however, to change the way that government uses socio-economic and financial evaluations, whether to optimise investment programming or pricing. Ensuring a coherent match between these two types of optimisation can provide a principle for determining the optimal programming price. We begin by showing that when projects are financed by both users (toll revenues) and taxpayers (subsidies), it is socially beneficial to plan these investments on the basis of the net present value (NPV) provided by each unit of public money invested. This NPV/subsidy ratio must obviously be higher than the public-funding scarcity coefficient or else the investment would destroy more wealth than it would produce. One of the ways of improving this ratio is also to optimise the toll level, since increasing it can lower the subsidy but has an adverse impact on the user surplus, it is essential to set the optimal toll. In the case of an approved project considered in isolation, we show that the optimal toll depends upon the public-funding scarcity coefficient. If there is no scarcity, the optimal toll is zero. As public-funding becomes scarcer, the optimal toll draws closer to the toll that optimises revenue. In the case of a programme of several projects subject to budget constraint, we show that the optimal toll no longer depends upon the public-funding scarcity coefficient and that there are several scenarios depending on the relative values of the maximum revenue and the total cost of the project: − when, whatever the toll, revenue can no longer cover over half of the cost, it is socially beneficial to choose not to levy any toll; − when there is a toll that covers the total cost, the operator may be left free to set it at the level he sees fit, with the issue of how the profits are to be shared between the franchisee and the franchisor being settled separately; − when the maximum revenue of the project falls between half and all of the total cost, the value of the toll that maximises the welfare function is lower than the revenue-maximising toll and must therefore be set for the private operator by government. Thus, the partnership contract must be given a different content in these three cases of optimal pricing. Preamble Most of the theoretical studies devoted to optimising public investment programming and infrastructure pricing have, since the work of Jules Dupuit (1844), focused on the salient issues of the transport sector, even though this work was relevant to all sorts of public infrastructure. The analysis presented in this paper deals with transport economics, but in line with this tradition, is also applicable to any field in which public-funding is combined with commercial revenue. For example, the question of determining how the financing of an opera should be shared between taxpayers and opera-goers raises the same type of issues as the optimal combination of tolls and subsidies for financing a motorway project. This report will be focused on these issues. We shall see that certain precautions are called for in investment programming together with some new thinking on pricing principles.