When Amber Post started grad school in physics at Princeton, her goal was the same as her male colleagues': a tenure-track job at a major university. Now with her Ph.D. just a year away, Post is thinking instead about working for a policymaking agency in Washington. Even though Princeton is generally welcoming to female scientists (the president, Shirley Tilghman, is a molecular biologist), Post, 25, senses that her reception in the larger academic world might be chillier. At elite universities, the percentage of women earning doctorates in science and engineering is considerably higher than the percentage of women professors--which means that a lot of talented women Ph.D.s like Post leave campus for jobs in government or industry instead of climbing the faculty ladder.

Stopping this female brain drain has been a challenge for years, and universities from MIT to Stanford are pushing hard with mentoring programs and stepped-up recruitment efforts. But Harvard president Lawrence Summers inadvertently threw the issue into the national spotlight when he suggested, at a recent academic conference, that women aren't succeeding because they lack "innate ability" in math and science. Scientists have uncovered some subtle differences in male and female brains, but it's unclear how these differences affect aptitude, and it certainly doesn't explain why there were only four women among 32 Harvard faculty members offered tenure last year. Around the country, women account for nearly half the bachelor's degrees in chemistry and math but only about 10 percent of the faculty. Summers's comments drew immediate criticism from women scientists at Harvard and elsewhere, and he issued a quick apology. "The human potential to excel in science is not somehow the province of one gender or another," he said in a statement.

But for many women scientists, the damage was done. "We all know that when we walk into a classroom or present research at a conference, there will be people who will think we are only there because we were given a break somewhere along the line," says Post. "But when the president of Harvard University appears to support the theory of innate differences, that pushes the stereotype into the realm of fact and makes it acceptable to think that women are just a little dumber by nature." MIT biology professor Nancy Hopkins, a Harvard alum who attended the Jan. 14 conference, walked out on Summers because, she says, his comments made her "sick." A few hours later, she told a reporter from The Boston Globe about Summers's remarks, which were made during a private session on women in science hosted by the National Bureau of Economics. After the Globe's story, Hopkins ended up on the "Today" show. That attention, she says, may be the only good to come out of the media firestorm. "People will realize what these women face," she says. "They must deal with men like Larry Summers.... They'll tell you they have no bias, but in their head they are thinking, 'Can women really do math?' "

In his talk, Summers proposed two other possible barriers for women: the conflict between tenure clocks and biological clocks, and outright bias against women (which he seemed to dismiss). Many women scientists blame these problems--not a lack of ability--for the dearth of women professors. Junior faculty need to spend their 20s and 30s on research and publication. Those are the same years when women have children. Time is an enemy for women in other professions, especially law and medicine. But while women doctors and lawyers benefit from lots of successful role models, academic science continues to be dominated by men. Harvard physics professor Melissa Franklin says that even though she has many talented women physics majors whom she tries to encourage, relatively few consider graduate school. "The atmosphere isn't compelling or welcoming to them," she says. "It's a very subtle thing."

And sometimes not so subtle. Lillian Pierce was Princeton's valedictorian when she graduated in 2002 and received her master's in math from Oxford as a Rhodes scholar. Now back at Princeton, she's studying for her Ph.D. Even with that sterling resume, she says, "I myself have experienced behavior that is hard to explain in terms of anything but discrimination: senior male mathematicians ignoring my presence when I'm introduced to them or suggesting point-blank that I pursue another career, such as medicine." She says too many of her female friends "drop out of graduate programs simply because they're disillusioned with the environment, not because they can't handle the math."

Even against this background, there has been some progress. Kristina Johnson, dean of engineering at Duke, says that more universities are making efforts to assist male and female faculty members with young families. "I've got 18 women on my faculty and the majority are raising children under 12," she says. "Yes, it's a challenge to raise kids and go through the tenure process, but it's one that both men and women face. Men don't want to miss out on raising their kids, either." But ultimately, the best remedy against bias would be more women on top, like Princeton's Tilghman, Susan Hockfield of MIT or Shirley Ann Jackson of Rensselaer Polytechnic Institute. "People need to see a career trajectory of someone who looks like them," says Jackson, "a trajectory that shows them that, with talent, a real career is possible." Although maybe not at Harvard--for now.