As technology and innovation seem to be contingent upon each other a great deal of attention has been given to the importance of assessing the contribution of R&D investment to firm and industry performance and, ultimately, to the economic performance of countries and regions. In industrialised societies not only private but also public agents have allocated increasing amounts of their resources to R&D activities, often considered the key path to innovativeness. At the same time, due to advances in empirical research, increasingly more focused on the micro (firms) rather than on the macro (country) level, old myths about the relationship between R&D, innovation and success began to fall down. Firstly, the idea that innovation is much broader than R&D has gained large support and has made it possible to identify other sources of innovation, beyond excellence in R&D, which had been largely hidden or neglected. As result, perceptions about small firms - or the so-called low-tech industries, which either do not carry out any significant R&D activities or are likely to perform them outside formal classifications - started to change. Secondly, the idea that more R&D investment is always automatically bond to success - whatever criteria one may choose to define success – has become nothing more than a utopia. In this paper we carry out an analysis of the literature on the magnitude and evaluation of R&D, and, possibly, of innovation. We identify the methodologies used and analyse to what extent the magnitude of (eventual) R&D returns is dependent on the methodology pursued and the level of analysis - firms (micro), industry (meso), and regions/countries (macro) - considered. We conclude that methodological approaches and levels of analysis determine, to a certain extent, the type of results obtained and, thus, variances between them.