A gas chromatographic method for the determination of cyanide residues in alcoholic beverages has been developed from procedures previously reported for application to water samples. Quantitatively isolating HCN from alcoholic beverages presented difficulties not encountered with aqueous solutions, particularly in the presence of S02 in the sample. HCN was liberated from the acidified sample by heating at 70 degrees C, flushed into an NaOH absorber solution, converted to cyanogen bromide (CNBr), extracted into diisopropyl ether, chromatographed on a Porapak Q column, and detected by an electron capture detector. S02 that is present in most wines interfered with the bromination step and caused low recoveries. This interference was eliminated by initially converting any cyanide present in the sample to the stable mercuric cyanide salt and then purging the acidified sample solution of all S02. The Hg(CN)2 present was then dissociated by adding KI and the analysis proceeded as previously described. Mean recoveries of 80-97% were obtained for 2-20 mug cyanide from replicate analyses of spiked samples of distilled liquors free of S02. The relative standard deviations ranged from 6.1 to 11.1%. Mean recoveries of 65 to 91% were obtained in the same range of cyanide from replicate analyses of spiked wine samples known to contain S02, with relative standard deviations ranging from 0.8 to 10.2%. The limit of detection was 0.2 mug HCN.