An alternative strategy is proposed for the disposal of spent nuclear fuel (SNF) and other forms of high-level waste (HLW) whereby the integrity of a mined and engineered repository for the bulk of the waste need be preserved for only a few thousand years. This is achieved by separating the particularly problematic components, notably heat generating radionuclides (HGRs) and very long lived radionuclides (VLLRs) from the waste prior to disposal. Such a solution requires a satisfactory means of disposing of the relatively minor amounts of HGRs and VLLRs removed from the waste. This could be by high-temperature very deep disposal (HTVDD) in boreholes in the continental crust [1,2]. However, the viability of HTVDD, and hence the key to the entire strategy, depends on whether sufficient melting of granite host rock can occur at suitable temperatures and whether the melt can be completely recrystallized. The high-temperature, high-pressure experiments reported here demonstrate that granite can be partially melted and completely recrystallized on a time scale of years, as opposed to millennia as widely believed. Furthermore, both can be achieved at temperatures and on a time scale appropriate to the disposal of packages of heat generating HLW. It is therefore concluded that the proposed strategy, which offers, environmental, safety and economic benefits, could be a viable option for a substantial proportion of HLWs.