AbstractPrevious research by the authors shows that the output and impact of British science has declined significantly over the last decade relative to other major industrial nations. This paper responds to the claim made by Leydesdorjf elsewhere in this issue of the journal that British research output has instead exhibited “a remarkable increase” from 1981 onwards. A detailed analysis of the available statistical evidence shows that Leydesdorffs claims are spurious, and serves as a caution to those attempting to use on-line bibliometric data-bases for science-policy purposes without first processing the ‘raw’ figures.