There is nothing more infuriating: you are an experienced scientist who has sent one of your best-ever papers to a journal, and what do you get back? A set of referees' comments that appals you. One reviewer asserts that the work is simply uninteresting and insufficiently original. Another displays wilful bias in relating their criticisms to results by a competitor whose outlook differs radically from yours. And a third has unreasonable expectations of what should be achieved. Not only are you upset, but your student co-author is devastated.

Such a trio of inadequate referees' reports would rightly make an author doubt the credibility of the editor, who has a duty to ensure fair play. But they might also give the offended author reason to reflect a little. What right has he or she to expect a high quality of peer review? What training is being given in his or her own lab to ensure that the next generation understands how to do a good job of critically appraising others' work? And as the pressures on researchers grow — bureaucracy from institutions and funding agencies, incentives to apply the outcomes of research — the very motivation to do a conscientious job of peer review is itself under pressure.

The fact is that the skills required to be a conscientious peer reviewer cannot and should not be taken for granted in young scientists. The culture of peer review is hugely variable — some heads of research groups take such training seriously, others do nothing. A more robust culture of good refereeing needs to be sustained — and that needs strong mentors.

Nature has for some years campaigned for good mentoring in the laboratory, particularly through a set of awards introduced in 2005 (see http://www.nature.com/npg_/community/nature_awards.html; the entry period for this year's awards, which are for researchers in South Africa, closes this week). On page 791, we publish a guide to all aspects of mentoring, derived from the diverse and very specific endorsements of good practice received through the nominations process.

Many of the goals of mentoring are self-evident. But the explicit fostering of ethics in the lab is rare. And ethical practice includes doing justice to other researchers in critically assessing their work.

So what are the elements of good peer review? The most important aspect is attitude, which should be one of constructive objectivity. This includes not only avoiding scientific bias, but also leaving behind any preconceptions about the labs in which the work was done.

The content should be constructive, too. To summarize the paper and highlight its strengths and essential significance not only shows a proper sense of engagement, but also helps both author and editor to benefit from a fresh perspective. But it is in making critical comment that a referee genuinely adds value, provided that such comments are constructively and collegially expressed. If there are experimental weaknesses or alternative interpretations, it helps to suggest experiments to strengthen the case or resolve ambiguities. If inadequate credit is given to previous work or if the paper fundamentally lacks originality, it is a good idea to provide appropriate references.

This may all seem obvious to the experienced reviewer. (Readers can find more on good peer review, and can comment on it and on this Editorial, at http://blogs.nature.com/peer-to-peer.) But it is only by careful oversight of a young scientist's attempts at reviewing real papers that the benefits of this experience can be passed on. Nature and Nature journals encourage the involvement of younger scientists by a referee, provided the authors' need for confidentiality is respected and the additional reviewer identified. Such mentoring should be routine in research team leadership, if standards are to be kept high.