There is a general belief that firms, one way or another, benefit from being located near other firms or organisations like universities or government agencies. This idea has attracted widespread interest among economic geographers, regional scientists and, more recently, policymakers. The problem is, however, that it is often merely assumed that these benefits exist and their existence is not actually tested. Therefore, there was a clear need to take a more systematic look at how and to what extent the performance of firms is affected by being located close to other organisations. To address that need, we have undertaken an empirical analysis of the Dutch software sector focusing on those questions. Unlike most previous studies on this topic, we have gathered data at the firm level to test some of the main assumptions regarding the benefits of spatial proximity. In all, we interviewed 265 software firms, a group that is representative of the large population of small software firms in the Netherlands. The first assumption that we tested is that firms located in regions where many other organisations are concentrated perform better than firms located outside such areas. Firms that are located in each others’ vicinity benefit from agglomeration economies and, as a result, are more likely to benefit from the spill-over of knowledge. Many studies that have tested this assumption have used data pertaining to the regional level and, therefore, could not control whether the effect may be caused by differences between the firms located in the various regions. Using data on the firm level, we could control for some firm characteristics (the age, size and innovation input of the firm and the way they dealt with changes in the market). We found that despite the widespread urbanisation and small size of the Netherlands still regional differences are large enough to affect the performance of software firms. A location in a region where computer services are concentrated positively affects the innovative productivity of ...