Monique Ferguson nearly slipped through the cracks. Though she was a top student in high school and college, she faced a bumpy road as an African-American woman pursuing a science career in what she felt was "a good-old-boys system." At the University of Texas Medical Branch (UTMB), where she attended graduate school, she was the first African-American to graduate from the department of microbiology and immunology. "It's hard when you don't have someone who looks like you in an administrative leadership position," she says.

Fortunately, the university's committee for diversity in graduate education reached out to her. The committee served graduate students in all fields, but luckily for Ferguson, the founder and chair happened to be a microbiologist. He was the only African-American scientist in the department, but one was enough. "He was the right person at the right place at the right time," she says. She believes his encouragement made all the difference. "I know for a fact that if he had not been my mentor, I would have pursued other options."

Today Ferguson is an assistant professor in the division of infectious diseases at UTMB. But for every Monique Ferguson who perseveres through the science "pipeline" and emerges successful at the other end, plenty of others are lost.

Women in the United States now earn half of science and engineering bachelor's degrees, according to the National Science Foundation, and 38% of science and engineering PhDs go to females, according to the Commission on Professionals in Science and Technology (CPST). But women represent just one quarter of the science and engineering workforce and, similarly, a quarter of all employed science and engineering PhD holders.

Underrepresented ethnic minorities fare worse. Together African-Americans, Hispanics, and Native Americans make up nearly 26% of the US population but, according to CPST, earn just nine percent of science and engineering doctorates. Kids from underrepresented groups often give up before they've even entered the pipeline, says Lino Gonzalez, a chemist at Genentech in South San Francisco. "They don't believe they can do it. It's not something that they've ever seen, or that anyone in their families has ever done."

Johanna Carmel Egan, an Indianapolis-based chemist and vice president of project management for Eli Lilly and Company, notes that women often feel isolated in graduate school without a robust support group of other women. Then "when you come into industry, you experience the same phenomenon to some extent," she says.

Minority groups also face family pressure, but in different ways. "There's a correlation between race and lower socioeconomic status," notes John Matsui, head of the University of California at Berkeley's Biology Scholars Program, which mentors minority students. Undergrads from cash-strapped families may need to turn down valuable lab experience for part-time jobs, while family members may pressure college graduates to take a job rather than take on debt for graduate school. "A lot is riding on what kind of job you get," Matsui says.

Even supportive families may not understand science or science careers. "My parents were laborers. They supported me, but couldn't sit down with me and teach me calculus," Gonzales says. Avery August, an associate professor of immunology at Penn State University, also had trouble explaining his career choices to family. "Our families often don't recognize what it takes to get a PhD, and they don't really understand the process," he says. "You have to explain why you're still in college after five years."

It's clear that many factors conspire to push underrepresented groups out of the pipeline, but the challenges aren't new. "There's been a lot of noise [about increasing diversity] for years, but there has been no systemic change," says Elizabeth Ivey, a retired physicist and past president of the Association for Women in Science. "That's the nut to crack."

Industry appears to be making some progress in this direction. Egan says diversity at Eli Lily has improved a lot in the 16 years since she joined the company. Similarly, Lino Gonzalez has positive feelings about the culture at Genentech, where 49% of employees are women and 43% are minorities. A big factor in that diversity success is the company's commitment to mentoring and networking. Genentech offers a variety of internal networking groups; Gonzalez belongs to one specifically for Latinos. That formal support structure helps to recruit and retain new hires. "It's a working group, we're working to achieve things, but it's also a support group," he says. "You have a lot of friends there."

Such structured programs are often the missing factor from diversity initiatives at colleges and universities, Ferguson says. She notes that the medical school at UTMB has a great diversity record.

According to the most recent figures available from UTMB, underrepresented minorities accounted for nearly 26% of graduates from the school of medicine between 2002 and 2005 - the same percentage as that of the underrepresented groups in the general population. Yet during the same period, minorities made up just nine percent of graduates from the school of biomedical sciences.

Ferguson attributes the difference to a program that pairs first-year med students with mentors who will guide them throughout their career. "It really seems to work, and the institution supports it," she says, "But those types of mentoring programs aren't built into the system at the graduate school."

Ferguson became a member of the committee for diversity in graduate education that once helped her through grad school, but the committee broke up due to lack of financial support, she says. "We have to have the support of the university to carry out our objectives." Unfortunately, Ferguson feels that support is often lacking - and not just at UTMB. Unless institutions are held accountable for increasing diversity, she says, nothing will change.

John Matsui, at Berkeley, couldn't agree more. He argues that many institutional diversity programs are working from the same untested checklist - without questioning the conventional wisdom, let alone challenging it. According to Matsui, "the big question is: What works, what doesn't, and for whom?"

Answering that question, he says, is where institutions and science faculty have failed miserably. Professors are good at telling students what to do, but terrible at listening. "We say 'we have this to offer' and if our students don't succeed, we place the blame on the students," he says. He argues it's time to turn the blame inward.

The irony, Matsui points out, is that scientists and science faculty have been taught to be skeptical and to hone their questioning skills. Yet "when it comes to diversity work, we take it at face value," says Matsui. He insists qualitative research must be done to really understand what works and why. Unfortunately, he says, many scientists don't respect social science enough to consider it. "There's been a reluctance to incorporate qualitative data into the design of diversity programs."

How to get past the reluctance? Carrots and sticks. Matsui argues that institutions could do more to reward faculty for efforts to improve diversity in their labs. The number of papers a professor publishes has a direct impact on tenure. But his or her efforts to improve education and increase diversity have little or no bearing on career advancement. "We respond to reward structures," says Matsui. "If our funders place outcomes front and center as requisites for funding, then we're going to hop to."

After all, it's not just about doing the right thing; it's about doing better science. Scientists ask questions based on their experiences, Matsui says, and broader backgrounds will only mean broader discovery. "What does diversity have to do with research?" he asks. "It has everything to do with it."