A picrite dyke with an olivine-bearing chilled margin and an olivine-rich centre has been used to test for the presence of a boundary layer around rapidly grown olivine crystals and for any variations in the olivine/melt partition coefficient for uranium as a result of probably different crystal growth rates. The technique of fission-track mapping is shown to be suitable for this kind of study despite the very low uranium concentrations in the olivines. A boundary layer appears to be present around some olivine crystals but it is not a consistent feature. Uranium partition between olivine and melt was not affected by different crystal growth rates, as revealed by different crystal morphologies.