Elias Zerhouni, director of the US National Institutes of Health, has come up with a simple way to improve the lot of postdocs. It doesn't involve bureaucracy or legislation or extra pay — just more communication and a refusal by postdocs to accept bad situations. “The best thing is if the brightest mentees walk away from places that don't have good mentoring,” Zerhouni told Naturejobs (for a complete transcript, see the Naturejobs graduate channel, http://www.nature.com/naturejobs/channels/graduate).

Of course, Zerhouni isn't suggesting that postdocs quit their labs en masse in fits of pique. Rather, he says that before postdocs accept their fellowships, they should sit down with their prospective principal investigator and negotiate their mutual expectations. They can then draw up a sort of scientific “prenuptial agreement”, Zerhouni says.

The understanding they reach should include agreement on how much career guidance the mentor will provide, as well as the amount of training that they can expect in such areas as presentation and grant-writing skills. Zerhouni adds that every postdoc will want something different from their mentor. Some prefer a hands-on approach with frequent reporting, whereas others may want a laissez-faire approach.

Zerhouni has a point. Postdocs have power in numbers. Their fellowships mark the last time in their career at which the supply-and-demand formula works in their favour. There are plenty of positions available, especially for postgrads who already have a good track record. Because postdocs need good publications and recommendations from a principal investigator to take the next step, it's essential that they find someone who will help them. And as Zerhouni says, if the principal investigator isn't willing to help, they should find another who will.