For an outsider, it's hard to imagine that conditions at Los Alamos National Laboratory in New Mexico could get much worse. When two disks containing classified data went missing last summer, federal investigators descended on the lab, operations were suspended for months, and scientists were berated as “buttheads” by Peter Nanos, the lab's director. Now, it turns out, those disks never existed (see Nature 433, 447; 2005 10.1038/433447a).

Incredibly, Los Alamos could slide still further downhill. Its long-standing contract with the University of California, which manages the lab for the US Department of Energy, will be opened up to competition. At stake for lab employees is the academic clout that comes with the university's reputation, as well as a sterling pension plan for which many at the ageing lab are already eligible. Unless these benefits are protected, insiders fear that scientists will flee in droves.

Since Los Alamos was founded in 1943, its relationship with the university has been good for scientists. They enjoy the same privileges as university employees, including a lower tuition rate to the university's campuses and membership of the retirement programme.

But since 1999, university officials have spent countless hours responding to scandals, so some wonder whether it is worth the $9-million fee that the university receives. And the energy department has been embarrassed again and again by a never-ending string of security incidents — many of which may have been inflated by political animus towards the laboratory and its managers.

Looking for others interested in running the lab isn't a bad idea, but the exercise is creating too many uncertainties for scientists. If the lab were to go to another company, the university's pension plan and other benefits could be lost for ever. Along with the director's penchant for name-calling, it is no wonder that many older scientists are reportedly taking early retirement.

The best hope for fixing the contract might be Samuel Bodman, the incoming energy secretary. Bodman's credentials as former professor of chemical engineering at the Massachusetts Institute of Technology played well with the lab's rank-and-file during a meeting in February. And his previous job as president of Fidelity Investments has won him praise from congressional overseers. If he chooses to, he may be able to rebuild the lab's broken contract with the university, restore trust between the wounded parties, and ensure that researchers' benefits are protected. If he doesn't, Los Alamos could become just another ghost town in the New Mexico desert.