Abstract: 1997 grey literature was defined at Luxembourg as "that which is produced on all levels of government, academics, business and industry in print and electronic formats, but which is not controlled by commercial publishers". Distinction between commercial (white), non-commercial (grey) and other (dirty) literature seemed easy. Six years later, the landscape of scientific and technical information has radically changed. What is the place of grey literature between continuous electronic resources, open archives and new economic models of scientific literature? Is there any place left? The authors describe the evolution and relative decline of traditional grey literature from a French perspective and present some actual trends in the processing of grey literature by INIST-CNRS.