This article criticises the analytical basis and the use of cost benefit analysis on several (though non exhaustive) grounds, but above all for its intrinsic inability to deal with problems of development; this depends on the assumption of marginality and independence of projects. Development and change require instead complementary, system-relevant actions, well coordinated among agents and through time. On such a basis, a link is established between the projects aimed at fostering a development process and the activity of development programming. By associating analytical considerations drawn from bounded rationality and empirical suggestions deriving from companies’ practices, several proposals are developed, concerning useful procedures for connecting sequentially projects and programs; such procedures - which exploit the experience made in the management of the European structural funds - allow not only a better coordination among involved agents, but to capture and use the learning phenomena that are generated by the processes of change and are at the roots of further development.