We studied the relative efficacy of different forms of foliar iron (Fe) fertilisation on leaf re-greening in Fe-deficient, purple-fleshed sweet potato (Ipomoea batatas (L.) Lam.) varieties xuzi8 and xuzi6. Activities of ferric chelate reductase (FCR) and concentrations of Fe were measured in the leaves and roots at intervals over 5 days to quantify recovery from leaf chlorosis. Freshly expanded and chlorotic leaves were immersed in one of three different fertiliser compounds containing 9 mm Fe: FeSO4, Fe2(SO4)3, Fe(III)-EDTA. An Fe-sufficient treatment and an Fe-deficient control were included. The experiment had a completely randomised block design with five replications per treatment and was conducted in a sunlit glasshouse. For variety xuzi8, leaf FCR activity in the Fe2(SO4)3 treatment was highest at 1 h after application, and higher than all other treatments, whereas FeSO4 and Fe(III)-EDTA treatments showed their highest FCR at day 5 after application, both significantly higher than the Fe2(SO4)3 and control treatments. Furthermore, leaf Fe concentration reached a maximum in the FeSO4 treatment at day 1, and in the Fe2(SO4)3 treatment at day 3. By contrast, root Fe concentration was relatively constant and lower in the foliar Fe treatments than the Fe-sufficient and -deficient treatments. For variety xuzi6, leaf SPAD was higher with the Fe2(SO4)3 than the FeSO4 treatment at day 5 after application. In general, FCR activity and Fe concentrations in roots and leaves of xuzi6 were higher than those of xuzi8. Variations in leaf Fe concentrations were similar for both the FeSO4 and Fe2(SO4)3 treatments of the two varieties. Maximum leaf Fe levels in xuzi6 were ~4-fold those in xuzi8. The results of the study suggest that foliar-applied Fe2(SO4)3 was the most effective compound at correcting Fe-deficiency symptoms. The higher leaf and root FCR activity and Fe concentration in xuzi6 might explain its higher tolerance to Fe deficiency and better re-greening than xuzi8.