A more quantitative extraction of arsenic-containing compounds from seafood matrices is essential in developing better dietary exposure estimates. More quantitative extraction often implies a more chemically aggressive set of extraction conditions. However, these conditions may result in undesirable chemical changes in the native arsenicals which may further complicate the toxicological risk assessment. This balance between quantitative extraction and species-specific integrity may be best addressed by using simulated gastric juice as an extraction solvent to mimic 'bioavailability'. This, conceptually, should extract the bioavailable fraction and induce any chemical changes that would occur because of ingestion. The most chemically labile species associated with seafood are thought to be the arsenosugars and for this reason their chemical stability is investigated in this study. Four arsenosugars (3-[5'-deoxy-5'-(dimethylarsinoyl)-beta-ribofuranosyloxy]-2-hydroxypropylene glycol, As(328); 3-[5'-deoxy-5'-(dimethylarsinoyl)-beta-ribofuranosyloxy]-2-hydroxypropanesulfonic acid, As(392); 3-[5'-deoxy-5'-(dimethylarsinoyl)-beta-ribofuranosyloxyl-2-hydroxypropyl hydrogen sulfate, As(408); and 3-[5'-deoxy-5'-(dimethylarsinoyl)-beta-ribofuranosyloxy]-2-hydroxypropyl-2,3-hydroxypropyl phosphate, As(482)) were isolated from seaweed extracts and subjected to simulated gastric juice and acidic conditions which mimic the stomach's pH of 1.1. Three acid solutions were used to test the chemical stability of the arsenosugars: simulated gastric juice, 78 mM nitric acid and 78 mM hydrochloric acid. The composition of the solutions was monitored over time (up to 48 h) using IC-ICP-MS for detection. The arsenosugars were found to degrade at the rate of 1.4% per h at 38 degrees C and 12.2% per h at 60 degrees C. The plots of percent conversion versus time were found to be independent of the starting arsenosugar and all had r2 values of greater than 0.97. A single common degradation product was observed in all the stability studies. A mass balance between the starting arsenosugar (As(392), As(408) and As(482)) and the degradation product was conducted with each set of experiments. This mass balance indicated that the degradation process did not produce any unchromatographable species. This degradation product was tentatively identified as As(254) as determined by ESI-MS/MS spectral data. An acid hydrolysis mechanism was proposed for the formation of As(254) from each of the native arsenosugars by hydrolysis at the C-1 carbon on the ribose ring.