Abstract Clusters of phosphoserine residues in casein bind iron with high affinity. Casein inhibits iron absorption in humans but partial hydrolysis of casein prior to ingestion diminishes this inhibition. The objective of this study was to test two hypotheses: 1. Partial hydrolysis of the peptide bonds in casein exposes phosphoserine residues to attack by intestinal alkaline phosphatase (IAP). 2. Hydrolysis of the phospho-ester linkage in phosphoserine residues in casein by IAP releases bound iron or inhibits iron chelation, thereby allowing its absorption. Test of hypothesis 1: Suspensions of sodium caseinate (SC), enzymatically hydrolyzed casein (EHC), and casein phosphopeptides (CPP) were subjected to an in vitro pepsin/pancreatin digestion and subsequently incubated in the presence of calf IAP. The rate of release of inorganic phosphate was measured with the following results (expressed as μmol phosphate released/unit of IAP/min): 0.081, 0.104, 0.139 for SC, EHC, and CPP, respectively. These results are consistent with hypothesis 1. Test of hypothesis 2: 59Fe-citrate or 59Fe-citrate + CPP in minimum essential media were spiked with a Na 2WO 4 solution or water (Na 2WO 4 is a known inhibitor of IAP) and placed on Caco-2 cell monolayers. Uptake of 59Fe by the cells was used as an index of iron bioavailability. Na 2WO 4 did not affect 59Fe uptake from samples containing only iron but did slightly inhibit (by 10%) uptake from samples containing iron + CPP. These results are consistent with hypothesis 2 and provide a possible explanation for the observation that partial hydrolysis of casein improves iron bioavailability.