The problem of the U–Pb age of zircon crystals from ultrabasic rocks is discussed in this paper. It is shown that the assumption on the xenogenic nature of zircon crystals in dunite is not consistent with the petrographic and experimental data. The results of experimental study of the zircon–baddeleyite transition and thermodynamics of the reaction of zircon replacement with baddeleyite show that these transformations are the likely way of the formation of zoned zircon crystals in dunite. Each zone of these crystals may have its own age.