Fourteen extraction methods commonly cited in the literature were evaluated for the quantitation of methionine (Met) and selenomethionine (SeMet) in a yeast candidate certified reference material (CRM). Species specific isotope dilution (ID) gas chromatography-mass spectrometry (GC-MS) was utilized to effectively compensate for potential errors, such as losses during derivatization and clean up steps. Despite different extraction methods, the same derivatization procedure using methyl chloroformate was applied with a single exception, which was based on digestion with cyanogen bromide with 2% SnCl2 in 0.1 M HCl. Significant differences in measured Met and SeMet concentrations were obtained when different extraction methods were used. A 4 M methanesulfonic acid reflux digestion was found to be the most efficient for both analytes. Digestion with CNBr with 2% SnCl2 in 0.1 M HCl for the determination of SeMet showed the second highest extraction efficiency. Despite frequent use of enzymatic hydrolysis for the extraction of SeMet from yeast, very low extraction efficiencies for both analytes were obtained for four of eight tested methods. Among these, the highest extraction efficiencies for both analytes were obtained using 20mg pronase and 10mg lipase with incubation at 37 degrees C for 24 h. However, recoveries remained nearly 30 and 50% lower for Met and SeMet, respectively, compared to extraction with methanesulfonic acid. Lowest extraction efficiencies for both analytes were obtained when HCl or tetramethylammonium hydroxide (TMAH) digestions were used. Efficient extraction was also achieved using 200 mg (or 400 mg) of protease XIV with incubation at 37 degrees C for 72 h (or 24 h). Concentrations of 3331+/-45 and 3334+/-39 microg g(-1) (mean and one standard deviation, n = 4) for SeMet were obtained using 200 mg (72 h incubation) and 400 mg (24 h incubation) of protease XIV, respectively, in agreement with a value of 3404+/-38 microg g(-1) obtained using a methanesulfonic acid reflux.