National economic growth is fueled by the development of high technology clusters such as Silicon Valley. The contributors examine the founding of ten clusters that have been successful at an early stage of growth in information technology. Their key finding is that the economics of starting a cluster is very different from the positive feedback loop that sustains an established cluster. While ‘nothing succeeds like success’ in an established cluster, far more difficult, risky and unlikely are the initial conditions that give rise to successful clusters. The contributors find regularities in the start of the successful clusters studied, including Silicon Valley 40 years ago. These cases contain ‘old economy’ factors such as competencies, firm building capabilities, managerial skills, and connection to markets, more than the flamboyant ‘new economy’ factors that have been highlighted in recent years.