The attack of 11 September 2001 changed many things, including the way science is done in America. Before the attack, the United States was a focal point for scientists from around the world who went there to study, conduct research and teach. But in its aftermath, a series of restrictive new immigration measures led to lengthy delays for foreign researchers, and many were turned away. In 2004, we warned that these policies were strangling scientific exchange and could have dire consequences for the nation's scientific leadership (see Nature 427, 181; 2004 10.1038/427181a).

As reported on page 6 of this issue, things have improved, although perhaps not as much as many researchers would have liked. The Department of State has boosted staffing levels at its embassies and consulates, and new computer systems are helping to prevent applications from becoming lost during interagency security reviews in Washington. Waiting times for those reviews are down from months to weeks: in Beijing, a student visa that might have taken six months can now be granted in just ten days.

The government has taken other positive steps, abandoning a plan to force foreign scientists to obtain licences to operate lab equipment (see Nature 441, 679; 2006), and revising a rule that would have required foreign-born researchers on defence-funded projects to work in 'segregated areas'. These moves were in response to a vocal and concerted campaign by a coalition of scientific, university and industry organizations that warned policy-makers that such restrictions would not be in the nation's best interest.

Researchers entering the United States today still find it a hassle, but most feel that it is worth the trouble to come to US laboratories. New statistics are showing a resurgence of foreign students, and researchers are probably following, although numbers reflecting their movement are more difficult to obtain.

This could lead some to believe that the problem is largely solved. But more should be done to ensure free scientific exchange between the United States and other countries. Scientific groups in America should continue to press for better training of embassy staff, and for the hiring of more scientifically aware case-workers. They should work with the State Department and the Department of Homeland Security to try to bring some much-needed transparency to the visa process. It is unlikely that the security checks at the heart of many researchers' delays will become more open, but more could be done to communicate with those entering the United States about the status of their applications. Finally, scientific organizations should support immigration legislation that would make it easier for foreign scientists trained in US universities to remain after finishing their work.

All this should be done with an eye to keeping an open dialogue between scientists and the federal officials overseeing immigration. In the event of another attack on the United States, such lines of communication could help to ameliorate the immigration restrictions that would almost certainly follow. The United States is a linchpin of the global scientific enterprise, and it is in everybody's interest that it remain open for business.