For about 20 years, a rather wide range of conceptual approaches to the social study of science and technology have emerged which have occasionally been la- belled "postconstructivist". Although these conceptions differ in various respects, they have in common a twofold opposition: against traditional representationalist realism as well as "classical" social constructivism established by the "sociology of scientific knowledge" (SSK). In order to escape the pitfalls of both these views (and to overcome the familiar, yet unfruitful opposition between them), postconstructiv- ist perspectives understand and study the sciences primarily in term of their situ- ated material and discursive practices . The present article starts with a brief retro- spect on why and how since the mid-1980s postconstructivist trends have ques- tioned not only rationalist and realist accounts but also the conceptual foundations and background assumptions of SSK's claim to explain sociologically the content of science. Subsequently, the central features of a postconstructivist perspective in science studies are outlined, referring to the key concepts of "knowledge", "prac- tice", and "performativity". The fruitfulness of a theoretical approach focusing on scientific practices is illustrated using the example of the increasingly important issue of scientific non-knowledge: In the same way that knowledge is not to be comprehended as simply the mental "possession" of a knower, non-knowledge is not merely the lack thereof but an (unrecognised) implication of materially and so- cially situated research practices. Finally, it is emphasised that postconstructivist science studies should not be misunderstood as claiming (as do realism and con- structivism) to provide a meta-theoretical explanation or legitimation of science. Instead, postconstructivism should be conceived as a situated critical effort to chal- lenge one-sided accounts of scientific knowledge and foster more self-reflective re- search practices.