A career in science requires planning — from drawing up grant proposals to seeking tenure. But, as for most things in life, events rarely follow the path laid out. This has readily been demonstrated over the past two years by the eight PhD students who have detailed their experiences in our Graduate Journal column.

Most PhD students intend to finish their studies in a set number of years, emerging with tangible results and, ideally, solid job opportunities. But is that usually the case? To find out, we have invited the four students who launched our Graduate Journal in 2004 to return this month for an alumni special.

The first to report back is Sidney Omelon, a PhD candidate at the University of Toronto, who is finding that writing a thesis can take longer and present more challenges than initially predicted (see page 240).

And the class of 2005, who wrapped up their contributions last month, have already shown how unpredictable life can be. Jason Underwood, for example, expected to complete his PhD last year at the University of California, Los Angeles, then settle down to a local job and life with his wife. By February, his wife had left him and his carefully laid plans had to be rethought (see Nature 433, 782; 2005).

Why should we make a note of such twists and turns? Because one running theme in Naturejobs is the need to evaluate risks. This is useful at all stages and skill levels of the scientific process.

But another reason is that Naturejobs is this year running a competition to find its next crop of PhD journal-keepers. Students interested in sharing their experiences should visit http://www.nature.com/naturejobs/magazine/competition.html — but be quick as entries must be received by 20 January. We can make no guarantees of what will come of this process, only that, if the past two years are anything to go by, it is bound to be interesting.