The growing role of knowledge as the base of the economy has meant growing expectations of universities all over the world to function as engines for regional growth. The independent role of universities is slowly being replaced by governmental policies for human capital formation, knowledge dispersion, innovation systems, triple helix, etc. One example is Sweden’s new University Act that added a third task to universities’ two traditional tasks, education and research,viz. cooperation with surrounding society. Theoretically, this change in policy is supported the hypothesis presented by Gibbons et al (1995) of an emerging Mode 2 of knowledge production. Based on Swedish, Scandinavian and international experience, this paper summarizes knowledge of regional effects of universities and higher education. One conclusion is that the “regiment effect” (Florax 1992) seems to be the most obvious regional effect of universities and that hopes for university-led innovative regional development have hitherto seldom been fulfilled. The paper also analyses the obstacles to more intimate cooperation between universities and surrounding society and knowledge production a la Mode 2. This analysis is performed by applying the concept of social capital. Two of the conclusions are that most regions do not have the capacity to absorb the output of the universities (Florida & Cohen 1999), and that the internal social capital of universities is not adapted to governments’ demands, nor are the relations between universities and other stakeholders in regions.