Funding pressures from governments have long helped to determine the direction of scientific research. But since the 1980s, the mantra 'publish or perish' has become increasingly pervasive. With more and more scientists competing for the limited space in the high-impact journals that institutions use to assess performance and award jobs or tenure, career progression seems an almost impossible challenge — especially for young scientists.

So, faced with such fierce competition, how does a postdoc gain a fellowship, or even a position in an institute? Grace Wong, chief scientific officer at ActoKine Therapeutics, a biotechnology company in Boston, says that apart from publishing in high-impact journals, young scientists can advance their careers by working with mentors who have strong international reputations ( Nature Biotechnol. 23, 265; 2005).

But working in a high-profile lab can brings its own problems. The level of competition can mean that younger scientists may have their work stamped on by senior colleagues. In the top laboratories, discoveries can lead to ugly situations, such as arguments over authorship or credit. And some principal investigators will assign two or three people to the same project to increase chances of success; the first to get the answer gets credit, a paper and perhaps a job. Reputation then becomes a very delicate issue and objectivity in determining scientific output becomes distracting.

Ideally, other measures of success could be used. One suggestion has senior scientists being evaluated not solely on the number of papers they have published, but on how many people's careers they have helped (see Nature, 434 801; 2005). If this approach were embraced, the need to establish senior people in institutes to advise postdocs and students on scientific issues would become greater.

Other possible changes include creating equality within the laboratory instead of a hierarchy, so that each postdoc and PhD student has equal access to the mentor's experience. And establishing clear demarcation for members of the same lab and between labs on a project could remove unnecessary competition, and allow everyone the chance to publish something.

Judging young scientists not just on publications but on other contributions would create a more welcome climate. But this won't happen unless scientific managers are rewarded for creating such an environment.