David Osterbur spent a decade pursuing an academic science career before tiring of the "never-ending cycle" of unfunded grant applications, he says. When his wife, like him a developmental biologist, accepted a job offer in Massachusetts, he took advantage of the change in location to weigh a change in career. He was considering a career in public health so he could continue using his science background, when his wife suggested he become a science librarian. "I had always enjoyed being in the library. In graduate school, people would always come to me when they couldn't find something," he says.
His Ph.D. in genetics and his research career studying Drosophila development made Osterbur a hot commodity on the science library job market. Immediately after finishing his master's in library science  at Simmons College in Boston, Osterbur took a job with a small branch library of DuPont Pharmaceuticals in Boston. He stayed there just over a year before moving, in 2000, to Harvard University's Biological Laboratories Library . Five years later, he landed his current position at Harvard's Countway Library of Medicine . "I really enjoy helping people and teaching," Osterbur says. "This is a career where I can keep my hand in the science without having to have my paycheck or the paycheck of a graduate student or postdoc depend on the next grant."
Largely employed by universities, science librarians shape the library's collection, teach scientists how to search for information, and even directly assist with certain informatics-heavy research projects. Hiring managers and applicants say that the job outlook is positive for science graduates or working scientists interested in library careers; open positions can remain unfilled for years, says Brian Gray, an engineering, math, and statistics reference librarian at the Kelvin Smith Library  of Case Western Reserve University in Cleveland, Ohio. "If they're willing to relocate," notes Gray, science graduates, even those without a library degree, "can pretty much write their job ticket."
John Meier worked briefly as an engineer before deciding that he didn't like the corporate environment. His interest in information management led him to explore, and eventually embrace, library science. He was pleasantly surprised to discover that academic librarians teach and assist in research. Furthermore, compared with Ph.D. researchers, librarians are generalists, which he considers a plus. Another bonus was that his employer, the Pennsylvania State University  in University Park, places its librarians on the tenure track, with the same expectations to publish, teach, and serve the academic community as it places on professors in academic departments.
Meier now specializes in math, statistics, and patents. About one-third of his workday is spent facing the public, answering questions from walkup patrons or researchers with appointments for help with, for example, bibliographic software such as EndNote or BibTeX. "It doesn't call on deep expertise usually," he says. When he's not answering queries, "a lot of it is budgets, ordering books, selecting journal subscriptions," he adds, so he talks to faculty members and analyzes the usage of the library's resources so he can prioritize them. He also visits classrooms--some weeks include several classroom visits--to teach students about library research tools.
Working alongside him, other librarians specialize in other science and engineering disciplines, and still others help with computer tech support or more general front-desk queries. Not all librarians are so specialized; at some institutions, a science librarian may be the only person with a scientific background. Osterbur, in contrast, supervises a staff of 16 at a library focused entirely on medical science.
What librarians learn quickly--and try to impart to scientists--is that information retrieval is not as straightforward as Internet search providers would have them think. "Many of these specialized databases, even PubMed , do not work well if you treat them like Google," Osterbur says. A science librarian's job is to master the specialized tools that can save a researcher time--and then teach researchers how to use them.
The skills used at academic libraries translate to industrial and business library settings, and some science librarians consult for smaller biotech start-ups, says Osterbur. Osterbur himself consults for a Harvard bioinformatics project, making regular use of his genetics training.
Librarians can stick to the reference librarian path or move onto the administrative path. Administrators spend more time deciding where to allocate a library's resources and less time facing the public, if at all, than do reference librarians, who tend to be more directly involved in research and answering questions from patrons. People usually decide which to pursue based on their personalities, Osterbur says. He adds that personality also plays a role in his hiring decisions: "I look for a good knowledge base" in the relevant discipline, Osterbur says, but "the other thing is a good service attitude."
Librarians spend a lot of time giving advice and teaching people one-on-one, so teaching experience is a plus, too. Gray's staff is "spending more and more time on instruction and showing people how to do research and how to use resources," so he chose one of his most recent hires because she had experience teaching chemistry to undergraduates. "We don't have to teach her how to teach because she's already done that," he says.
Teaching experience may be why Christie Peters, a recent graduate of the University of North Carolina (UNC), Greensboro, library program , landed a job interview this month at a Texas library "without even a phone interview beforehand," she marvels. Peters got a bachelor's degree in biology and worked as a lab technician and as a high school teacher before returning for a second bachelor's in the history of science as preparation for a research career as a historian. But she decided she wanted to stay closer to science, so she enrolled in library school instead. While working on her master's degree, Peters sought out faculty members in the chemistry department at UNC Greensborough and helped them integrate library research methods into their undergraduate courses. "What really seems to get me the interviews is that I've worked with chemistry faculty," she says.
Perhaps the best way for scientists to decide if they might like a career in the library instead of in the lab is to work part-time in a library. In addition to her faculty outreach, Peters has interned at three different libraries. "This is a much better fit for me" than a traditional research career, Peters says: "I feel like I'm much more of a generalist."
Indeed, the 30,000-foot view of your discipline will be useful as a science librarian. For example, knowing the specialized genome or protein databases used in your lab will be useful as a science librarian, but you might also want to step back and learn about similar databases in related fields.
A library science or information science master's degree--which usually takes 1 year, full-time--is not always necessary for candidates with graduate work or an advanced degree in the sciences. But many library job advertisements do call for an accredited library science or information science degree, and Osterbur acknowledges that library school "really does teach you how to search." He and other science librarians also say that having a library degree can make it easier to integrate with other librarians.
And finally, although science librarians are in a unique position to take part in many different research projects, it's important to remember that "it's really a service job," says Osterbur: "You're not out to make a reputation for yourself."
Text corrected: 30 March 2009
Lucas Laursen  is a freelance science writer in Cambridge, U.K.