Nuclear power has performed badly in recent years as a new investment everywhere except Japan and Korea. This has mainly been for orthodox financial and economic reasons. Among the factors contributing to this loss of competitiveness, persistently rising real capital costs have been particularly important. While the nuclear industry has believed it could control and reduce capital costs, increasing regulatory stringency has made designs more complex and correspondingly more costly. These cost increasing factors have far outweighed traditional cost reducing factors (like learning). The only lasting way to meet increasing stringency in safety at acceptably low cost is likely to be the development of new and simpler reactor designs.