We assess the fragmentation of strange quark matter in astrophysical events, showing that the application of statistical multifragmentation models suggests that most of the fragments (strangelets) should belong to an unstable baryon number $A$. While there are some caveats to be addressed, the flux of strangelets onto the Earth could be orders of magnitude lower than previous estimates, with negative prospects for the ongoing search experiments.