For a long time it has been considered that knowing is classification, but correct classification. And, in fact, classifications are a key basis for any statistical activity. But how are the different branches in this activity classified? In these branches agricultural statistics are certainly different. But very often they are even considered as a separate case, beyond the specificities deriving from the special characteristics of their object, the agricultural activity. The article addresses what its contours are, how they are developing and asks: in what sense are they or should they be considered a separate case, distinguished not by the differences of the object, its field of knowledge, but also by an epistemological specificity (structure, methods and validity of that knowledge)?

Accordingly, the paper discusses the place of official statistics in agriculture; the presumed historic isolation of this sector's statistics; further demands on them; their specificities but also similarities with other sectoral socioeconomic statistics. Lastly, it addresses their interactions and (necessary) integration with other statistical domains.

Keywords: Agricultural Statistics; Agriculture

Biography: Agronomist from the University of São Paulo - USP (1980) and D.Sc. in Economy from State University of Campinas - UNICAMP (2006). He is currently Coordinator of Agriculture of the Brazilian Institute of Geography and Statistics. He has experience in economics, with emphasis on social and economic statistics, acting on the following topics: agricultural statistics, economic statistics, agribusiness, agriculture and sustainable development.