A newly released 'roadmap' for Europe's future research infrastructure is, first, a reminder that the continent already has quite a lot of it. The list ranges from the Large Hadron Collider at CERN, the particle-physics laboratory near Geneva, Switzerland, to the five campuses of the European Molecular Biology Laboratory, to the Isaac Newton Group of Telescopes in the Canary Islands near Spain. Such facilities open up research opportunities for scientists across the continent — and are particularly important for tapping the talents of scientists in smaller countries, who generally have less access to the kind of national facilities built by richer nations such as the United Kingdom, France or Germany.

Unfortunately, the roadmap is also a reminder that Europe still does not have any systematic way to plan, prioritize or fund these infrastructure projects. Each joint facility to date has been an ad hoc effort, with scientists often working for years to forge a coalition of nations willing to pay for it.

The resulting delays can be costly or even fatal. Biologists warn that life-science infrastructures, such as archives or access to distributed specialist resources, have not kept pace with the progress of the field (see Nature 447, 377–378; 2007). And Earth scientists are beginning to worry about the slow progress in finding partners for Aurora Borealis, a planned €635-million (US$870-million) research icebreaker ship, the technical design of which was finalized earlier this month.

The governments of the European Union are aware of the problem. In 2002, they set up the European Strategy Forum on Research Infrastructures (ESFRI), a group of senior science administrators who advise national governments and the European Commission on infrastructure needs. In 2006, the ESFRI released its first roadmap: a list of 35 already-proposed infrastructure projects that the forum deemed to be of pan-European interest. The projects on that list ranged from the large — a €1-billion, 5-megawatt neutron spectroscopy facility known as the European Spallation Source — to the small: a €12-million digital infrastructure for the arts and humanities.

Then on 9 December, at its meeting in Versailles, France, the ESFRI released an updated roadmap that lists 44 projects, including all but one of the original 35. The construction cost for all 44 initiatives over the next 12 years would be about €18 billion. But the EU's Seventh Framework research programme allocates just €1.7 billion to such costs; the shortfall will need to come from elsewhere.

The creation of the ESFRI is an important first step, but it is not nearly enough: the forum neither funds the projects nor sets explicit priorities among them. Some research communities have started to set their own priorities — a notable recent example being the astronomers (see Nature 456, 427; 2008) — and other communities should certainly follow their lead. But no one is setting Europe's infrastructure priorities across science as a whole.

Such priority-setting is only going to become more urgent as budgets tighten during the economic downturn. So the EU's next step should be to set up an independent European authority that has the power to evaluate infrastructure projects on the basis of their scientific promise, to prioritize them and, ultimately, to fund them. The European Research Council (ERC), which funds basic research on the sole basis of scientific excellence, could serve as a model.

Indeed, such a 'European Research Infrastructure Council' might even serve to stimulate additional funding — just as the ERC process has encouraged member states to fund top-notch projects that have been rejected by the ERC on the basis of its limited budget.

Granted, the creation of such a body would require a transfer of national responsibility and substantial research money — things that are hard to make palatable to national governments. And the new organization would have to deal with the headaches that follow from Europe's wildly varying finance, licensing, tax and social-security systems. Nonetheless, it could pool money, competence and negotiating power, and vastly improve the painstakingly slow process of supplying scientists in Europe with the tools they need to produce the kind of collaborative research they are expected to deliver.

Scientists prove every day that true collaboration can be every bit as successful in Europe it is elsewhere. It is the responsibility of EU policy-makers to make sure that this doesn't change.