For around 250 British scientists, April's post will bring a particularly personal letter of rejection. The Engineering and Physical Sciences Research Council (EPSRC), a national research-funding body, has decided to stop serially unsuccessful applicants from submitting any more grant proposals for a year (see page 391).

No other funding body, whether in the United Kingdom, the rest of Europe or the United States has attempted to formally exclude scientists in this way. The anger triggered by the move tells a cautionary tale: the policy is misguided, however urgently the EPSRC needs to relieve its overburdened peer-review system.

Any benefit an applicant ban might have provided is being outweighed by bad feeling.

In truth, the research council is caught in a vicious circle. Its budget for grants shrank this year as government funding increases were swallowed up to pay for the full economic costs of research. Yet the applications keep flooding in. The result is that only one in every four or five applications is currently being funded; in some disciplines, such as chemistry, success rates have dropped to 15%. In the hope of getting at least some support, researchers feel compelled to submit more and more applications for small, short-term grants, further pressurizing an over-burdened system. At the same time, the plummeting success rates lead referees and applicants alike to focus on safe, incremental research rather than larger, more ambitious work.

But a ban on researchers — even those whose consistent lack of success disproportionately overburdens the system — is a clumsy way to try to break this cycle. When success rates are so low, the peer-review system cannot reliably identify the worst performers: rankings can vary so much from one reviewer to the next that many solid proposals end up being rejected along with the weak ones, just by the luck of the draw. And even if the system were reliable, the scientists involved have no time to adjust: the policy is being applied retrospectively. Worse still, this temporary ban could easily leave a permanent stain — particularly on the careers of young researchers.

Any benefit an applicant ban might have provided is being outweighed by the bad feeling its abrupt introduction has engendered. Some potential reviewers are talking of boycotts because they don't want to contribute to a system in which their decisions can shut out colleagues. Researchers also feel that they were not sufficiently consulted on the specifics of such a controversial decision.

Other options were available. Using an expert-committee triage to sift through outline applications before the full peer review, for example, would give instant feedback to researchers on where they are going wrong, speeding up the recovery process. Introducing regular deadlines for submission might help regulate the flow of applications from particularly voluminous applicants.

On its own, the ban is likely to achieve little and provoke much. It might have been more happily accommodated if accompanied by an overarching set of reforms, discussed with the wider community. Maintaining the peer-review system for grant applications depends on the trust and cooperation of its reviewers — the researchers themselves, who do the bulk of the work. The EPSRC seems to be alienating the very scientists its system depends on.