A new study has cast further doubt on controversial claims made earlier this year that nuclear fusion was achieved in a bench-top ‘sonoluminescence’ experiment. In the first measurements of their kind, Yuri Didenko and Kenneth Suslick of the University of Illinois in the US tracked the processes that occurred in a single bubble in water when it was compressed by pulses of sound. They conclude that endothermic chemical reactions would make it “exceedingly difficult” to reach the high temperatures needed to spark nuclear fusion in such bubbles (Y Didenko and K Suslick 2002 Nature 418 394).