Sir

Prompted by the report of a large difference between award rates for women and men in the postdoctoral fellowships competition of the Swedish Medical Research Council (Nature 387, 341; 1997), the Medical Research Council of Canada has assembled data on approval rates in its own programmes.

MRC-Canada expects that women and men should have equal opportunity in competitions for grants and awards. To that end, we advise members of selection committees that, when assessing the scientific achievements of applicants, they should take into account factors such as time devoted to child-bearing and raising.

For the Operating Grants programme, MRC-Canada's principal mechanism for supporting high-quality research projects, the approval rate over a three-year period was 25.4% for applications led by women (297 out of 1,167) and 26.6% for applications led by men (1,160/4,368). A Chi-square test reveals that the difference is not statistically significant.

In competitions for MRC-Canada scholarships, an award that provides five years of salary support for recently trained researchers, there is similarly no statistically significant difference between the approval rates for applications from women (14% or 20/143) and men (16.6% or 64/386).

MRC-Canada fellowships offer personal support for two types of developing researcher: PhD graduates who are pursuing postdoctoral training and health professionals who are undertaking intensive training in research. For five fellowship competitions, the overall approval rate is 12.9% for applications from women (111/858) and 16.3% for applications from men (222/1,361).

The difference is statistically significant (p < 0.05). Nevertheless, in two of the five competitions, the approval rates for applications from women and men are virtually identical.

We should expect to find variation in scientific review processes, not only among agencies around the world, but also among programmes. For example, at MRC-Canada we use classic peer review in the assessment of proposals for operating grants whereas for fellowships we ask a multidisciplinary committee to assess the research potential of candidates. It is reassuring that award rates in operating grants competitions indicate equal opportunity for women and men scientists. For research training programmes, such as fellowships, where candidates are not yet established scientists, we believe that assessment criteria should include not only the candidate's research accomplishments and projects (both of which may have been strongly influenced by past and present research supervisors) but also the candidate's critical ability, independence, perseverance and so forth.

We are working to ensure that none of the criteria involved is systematically influenced by gender.