Fe absorption from water-soluble forms of Fe is inversely proportional to Fe status in humans. Whether this is true for poorly soluble Fe compounds is uncertain. Our objectives were therefore (1) to compare the up-regulation of Fe absorption at low Fe status from ferrous sulphate (FS) and ferric pyrophosphate (FPP) and (2) to compare the efficacy of FS with FPP in a fortification trial to increase body Fe stores in Fe-deficient children v. Fe-sufficient children. Using stable isotopes in test meals in young women (n 49) selected for low and high Fe status, we compared the absorption of FPP with FS. We analysed data from previous efficacy trials in children (n 258) to determine whether Fe status at baseline predicted response to FS v. FPP as salt fortificants. Plasma ferritin was a strong negative predictor of Fe bioavailability from FS (P < 0·0001) but not from FPP. In the efficacy trials, body Fe at baseline was a negative predictor of the change in body Fe for both FPP and FS, but the effect was significantly greater with FS (P < 0·01). Because Fe deficiency up-regulates Fe absorption from FS but not from FPP, food fortification with FS may have relatively greater impact in Fe-deficient children. Thus, more soluble Fe compounds not only demonstrate better overall absorption and can be used at lower fortification levels, but they also have the added advantage that, because their absorption is up-regulated in Fe deficiency, they innately 'target' Fe-deficient individuals in a population.