Positivism has been declared dead in sociological theory circles, yet questions remain as to its viability among researchers. The authors present diagnostic evidence about positivism in sociological practice through a content analysis of journal articles published in the late 1960s and the late 1980s in the sociological journals of the USA and Britain. Using an index based on seven elements of positivism that were characteristic of the 'theory construction' movement of the late 1960s, the authors find evidence of the effects of time and nation on the use of positivism. Disaggregation of the index reveals that most of the observed change is associated with the elements of 'instrumental' positivism, particularly statistics. The results raise questions about the relationship between theory and research and about sociologists' philosophies of science.