Earthquakes are unpredictable. The quake that struck the Italian town of L'Aquila on 6 April 2009 and killed 308 people was no exception. Indeed, only days before the event, a group of experts that advises the Italian Civil Protection Agency had concluded that a large earthquake in the area was unlikely, but could not be ruled out (Nature 465, 992; 2010). Despite the carefully placed caveats of their assessment of the situation, the six scientists on the committee are now being investigated on charges of manslaughter, because they did not call for people to be evacuated.

The case poignantly illustrates the difficulties of communicating uncertainty in science. The scientists had clearly stated at the meeting that a large earthquake was still possible. But their protestations were not taken seriously enough by the deputy technical head of the agency, who summarized the committee's conclusions at a press conference with the words: “The scientific community tells us there is no danger, because there is an ongoing discharge of energy. The situation looks favourable.” A group of local citizens then filed a formal request for an investigation, claiming that many of the victims had planned to evacuate their homes, but changed their minds based on the committee's findings. The combination of summary, simplification and probably a degree of wishful thinking had led to a complete breakdown in the chain of communication between scientists and the public.

The outcome is disastrous on all fronts: people may have died or been injured unnecessarily; scientists are under charge of manslaughter for not predicting an event that is known to be perfectly unpredictable; and public trust in science has taken yet another battering.

The L'Aquila prosecutor's office will assess whether the call for an investigation has any merit. In the meantime, the scientific community is rushing to the aid of the scientists — for example, in a statement issued by the International Union of Geodesy and Geophysics (http://go.nature.com/qLKfu5). It is to be hoped that the seismologists of the Italian Major Risks Committee will not be prosecuted for giving imperfect advice to the best of their knowledge on an intractable problem. But, just as for El Niño predictions (Nature Geosci. 3, 231–232; 2010), this case illustrates the need for a closer analysis of the way in which scientific advice is transformed into actions by politicians and members of the public. Decision theorists may be able to help avoid catastrophes like the L'Aquila miscommunication in the future.