'Micro' sociology involves a focus on observable, day-to-day, face-to-face interaction; 'macro' sociology involves relationships one might call 'historical' rather than day-to-day, and spatially 'spread out. This paper discuss the relationship between the two as perspectives on the sociology of science and technology. It is argued that to see them as opposed is to miss a central aspect of the development of science, mathematics and technology. The shift, always potentially problematic, from the local (and thus from the terrain of 'microsociology') to the non-local is at the heart of what these are as human endeavours. So a satisfactory approach to understanding them must be both 'micro' and 'macro'. Though traditional 'macrosociology' is typically defended as being more relevant politically than 'microsociology', the former is insufficiently puzzled by the phenomenon of structural power. The concerns of the latter have potentially great relevance to an understanding of power, for example in an investigation of how scientific or technical change is made 'available' for explicit management, by capitalists, democratic representatives, or whoever. These abstract arguments are fleshed out by a discussion of one particular example: the construction of nuclear missile accuracy figures.