The significance of innovation for economic development is very little debated today. Instead, debated are the characteristics and preconditions of innovation. In regional studies and regional science, several approaches and theories have evolved in the last decades to explore the spatial dimension of industrial innovation activities. These approaches, however, have their origins in much older observations dealing with, for example, geographical agglomeration of production and its effects or inter-organizational interaction and its dynamics in space. Serving as the subject and frame of analysis in the study, the regional innovation system (RIS) approach represents the present theorization on relationships between industrial innovation and regional development. This approach is often counted as one of the variants in the innovation system family, together with national, sector and technological variants. It is noteworthy that these models have had a remarkable impact on policy-making all over the world in the last ten years, underscoring also their social attractiveness. A core assumption in the regional innovation system approach is that innovation activities are based on interactive learning and are thus socially embedded processes which cannot be understood without taking into account their institutional and cultural contexts. A broad definition of an innovation system formulates that an innovation system is formed by actors, relationships and processes related to producing, distributing and using economically useful knowledge. From the regional perspective, it is usual to emphasize those advantages for innovation processes that are caused by co-localization or close geographical proximity among actors (e.g. technological spillovers, specialized resources and institutions). Those characteristics of tacit knowledge are also emphasized that make it highly localized and “sticky” contrary to codified information that may be relatively easy and cheap to transmit to another part of the globe by exploiting communication technologies, for example. One of the points of departure in this study was an assumption that a study conducted in the RIS fashion typically follows a topdown methodology, although also studies from the bottom-up perspective are needed in order to facilitate comparisons between different empirical studies and to ensure an effective cumulation of research knowledge. This is where the study aims to make its mark: based on the RIS approach, it pursues to develop a framework that takes into account firm-level core determinants of innovation activities and its spatial ties. Here firms’ external interaction related to innovation as network-type linkages to other firms and organizations as well as its potential concentration on the regional level are considered to be of particular interest. Empirically the study is based on two connected firm-level data-sets, one of which consists of a survey with 360 respondent firms and business units and the other of 35 firms based on 38 executive interviews. These firms represent manufacturing and business services industries and they are located in Tampere and Jyväskylä Regions in Finland. The regions, on the one hand, are comprised of institutionally fairly rich innovation environments but, on the other hand, include also some less-favoured sub-regions. Concerning the analysis of the study, a combination of the two data-sets enables methodical triangulation which aims at securing both the diversity and the reliability of the outcomes. In addition to the constructed framework, the study results in empirical information on the Finnish firms’ innovation activities and their regional, national and international linkages. The outcomes show, among other things, that there are no particular innovation-related systemic inter-firm entities formed at the regional level. Based on this, the conclusion is that innovation systems, or environments, are spatially multi-level. On the other hand, the study brings forward many of those regional-level ties and linkages that are important for innovation activities, such as regional labour markets, universities and other science and technology infrastructure in the regions. Based on the study, many prospects are pointed out and discussed to contribute to the theoretical and methodological advance of the regional innovation system approach. Finally, originating from the empirical observations, three normative models of distinctive regional innovation environments are presented.
