In applying a set of standard bibliometric indicators to rank the scientific status of 500 universities worldwide for the 2010 Leiden Ranking, we have discovered that the language of publication has a dramatic and largely underestimated effect on citation-based measurements of research performance.

Publications in non-English-language journals count as part of a country's output, but these generally have a low impact as fewer scientists can read them. This effect is particularly evident in application-oriented fields such as clinical medicine and engineering, and in the social sciences and the humanities.

As clinical medicine represents a considerable part of the scientific output of most nations, the language of publication directly affects the ranking of the university hosting the research (T. N. van Leeuwen et al. Scientometrics 51, 335–346; 2001).

Rankings based on the number of citations per paper and per staff member are responsible for the strikingly low position of many German and French universities, particularly those that include a medical school.

Important ratings by Times Higher Education, QS, the Shanghai Academic Ranking of World Universities and the Leiden Ranking, for example, all unfortunately rely on rankings influenced by this language effect.