In order to resolve questions frequently raised in the context of rese
arch evaluation about the citation rates of journal publications in re
lation to other types of publications, the total research output of su
bstantial institutions or systems has to be brought under bibliographi
c control. That precondition has rarely been met: there are few publis
hed studies of the total range of publications of major research insti
tutions, including books, book chapters, technical reports and publish
ed conference proceedings. The Research Evaluation and Policy Project
(REPP) at the Australian National University (ANU) has established a d
atabase covering all the publications from the Institute of Advanced S
tudies (IAS), a fulltime research institution at the ANU, and has exam
ined in detail citations in the journal literature accruing to all typ
es of publications. The database contains a significant number of publ
ications, nearly 30 000 items, and covers the sciences and the social
sciences and humanities. This data enables us to examine whether the c
itation record of research publications appearing in journals indexed
by the Institute for Scientific Information (ISI) is a useable surroga
te for the citation record within ISI journals of all model of publica
tion. We contend that, if certain preconditions are met, the choice of
citation rare is not critical.