Chemical and enzymatic reagents have been employed to determine available concentrations of Fe, Mn, Cu and Zn in contaminated estuarine sediment. Gastric and intestinal enzymes (pepsin, pH 2, and trypsin, pH 7·6, respectively) removed significantly more metal than was water-soluble or exchangeable (by seawater or ammonium acetate), while gastro-intestinal fluid of the demersal teleost, Pleuronectes platessa L. (plaice), employed to operationally define a bioavailable fraction of contaminants, generally solubilized more metal than the model enzymes. Manganese was considerably more available than Fe under these conditions and it is suggested that the principal mechanism of contaminant release is via surface complexation and reductive solubilization of Mn oxides, a process which is enhanced under conditions of low pH. Of the chemical reagents tested, acetic acid best represents the fraction of Mn (as well as Cu and Zn) which is available under gastro-intestinal conditions, suggesting that the reducing tendency of acetate is similar to that of the ligands encountered in the natural digestive environment. Although the precise enzymatic and non-enzymatic composition of plaice gastro-intestinal fluid may be different to that encountered in more representative, filter-feeding or burrowing organisms, a general implication of this study is that contaminants associated with Mn oxides are significantly more bioavailable than those associated with Fe oxides, and that contaminant bioavailability may be largely dictated by the oxidic composition of contaminated sediment.