There are many obstacles to diversity in science. In any nation, there will be cultural and societal factors — often intersecting — that prevent the full research potential of one population group or another being fulfilled. One manifestation is discussed on page 211.

We at Nature have attempted to put our own house in order, and have produced just a scratch on the surface of one particular challenge — the low proportion of women contributing to our own content. That scratch is there thanks to actions taken since we focused on this issue in an Editorial a little over a year ago (see Nature 491, 495; 2012).

So what have we achieved? In the visibility of women in our pages, progress has indeed been made.

In the News & Views section, the proportion of female authors has increased from 12% in 2011 to 19% in 2013.

The proportion of women appearing in profiles by our journalists has increased from 18% in 2011 to 40% in 2013. That does not include the four profiles in our ‘Women in science’ special issue early this year (see nature.com/women).

The number of articles by women in our World View section, which is driven by current topics, has remained low, now running at 12%. By contrast in 2013, 33% of Comment articles had at least one female author (27% of them had a woman as the first author). The combined total of World View and Comment articles with at least one woman author in 2013 is 26% — an improvement on 19% in 2011–12.

In our Editorial a year ago, we highlighted the need for a ‘gender loop’ — a conscious step in which an editor deliberately identifies several female candidates before selecting authors and profile subjects in our magazine sections, and referees for our research papers.

In this last category, the result has been disappointing — the number of women referees has remained all too low. From 14% in 2011, the proportion of women fell to 12% in 2012 and then rose to 13% in 2013. Taking into account uncertainties resulting from ambiguities in some names, these numbers are essentially on a plateau.

Efforts have been made by research editors, when visiting labs and meetings, and when surveying the literature, to increase the number of women invited to act as reviewers. Women already make up only a small proportion of the potential referees, owing to the demographics of the research community. And our efforts have made us all the more aware that a higher proportion of women than men decline our invitations to referee. We have not investigated this with a survey, but informal comments indicate that women tend to be that much more busy.

The lesson in this tale is: we must try harder.