We invite scientists to comment on our proposed tool for rating the impact of research publications on society (http://www.societalimpact.info). We believe that such ratings should eventually be incorporated into assessments of scientific performance.

This tool has been devised by a task force recruited from different disciplines at the Medical University of Vienna (G. Watts Br. Med. J. 338, b553; 2009). It evaluates factors that include the aim of the published investigation; the extent to which authors attempt to translate their scientific findings into societal action; and the level, status and target group of this translation.

Authors' efforts to convey the significance of their results for daily life are particularly relevant. The level of translation (regional, national or international), its status (preliminary or permanent) and its target group (individuals, population subgroups or the public) are other important quantitative indicators.

There will be challenges in implementing our system. Not least is the problem of measuring the likely societal impact of publications from the basic sciences. There will also be an inherent subjectivity in the assessment, and societal rating may not lend itself to computerized calculation.