The agronomic effectiveness of partially acidulated phosphate rock (PAPR) fertilizers depends on the proportion of water soluble P present and the long-term release of P from the unacidulated residual phosphate rock (water-insoluble residue) in the soil environment. The relative reactivity of water-insoluble residues in a range of partially acidulated phosphate rock (PAPR) fertilizers was assessed according to chemical solubility, and was found to be lower than the corresponding original PRs. Two causes were identified for the reduced formic acid solubility of water-insoluble residues: (i) some preferential dissolution of more reactive apatite components (i.e. higher carbonate substituted apatite) during partial acidulation as indicated by reductions in the molar apatite C:P ratio of the residual PRs compared with the corresponding original PRs, and (ii) acid-insoluble metal (Fe-Al-Mg)phosphate and/or fluoride complexes formed on the surface of residual PRs during partial acidulation with commercial-grade phosphoric acid, which were partially soluble in alkaline ammonium citrate. These two mechanisms accounted for most of the solubility differences observed for 'medium' reactive PRs (Khouribga, Arad), compared with only half for more reactive PRs (North Carolina, Gafsa). The conclusion from this study is that partial acidulation may not be an ideal method for the manufacture of fertilizers containing fast and slow release (i.e. PR) components, particularly when highly reactive PRs such as North Carolina are used.