Cluster Impact Fusion is a suggested method of producing practical fusion power using small clusters of heavy water molecules directly accelerated into a titanium-deuteride target. Calculations suggested that such a system enhanced the cross section by many orders of magnitude. It is a particular implementation of the larger beam-target fusion concept.

The idea was first reported by researchers at Brookhaven in 1989.^[1][2] Intrigued by recent reports of cold fusion, they attempted to study potential causes for the effect by accelerating tiny droplets of heavy water, about 25 to 1300 D₂O molecules each, into a target at about 220 eV. To their surprise they immediately saw fusion effects, at a rate that was many times what any of them could explain via conventional theory.^[3]

The experiment was fairly simple in concept but required an appropriate accelerator, so it was some time before other labs were able to repeat the experiments.^[4] One of the first was the University of Washington, who reported a null result in 1991. Further experiments and a review from MIT in 1992 solved the mystery: the fusion products were the results of contamination, which could be eliminated by filtering with a magnet.^[5][6][7] The Brookhaven experimenters tried this and the effect disappeared.

Cluster impact fusion references end abruptly at that point.^{[citation needed]}