Linamarin and lotaustralin were isolated as a mixture from Trifolium repens L. and used as substrate for the assay of linamarase. IN A STUDY of the genetic control of linamarinase activity in Trifolium repens L., it became necessary to isolate the native substrates, linamarin (α-hydroxyisobutyronitrile-β-D-glucoside) and lotaustralin (α-hydroxy-α-methylbutyronitrile-β-D-glucoside) for use in a quantitative assay of linamarase activity in crude leaf extracts. The first report of a preparative scale isolation of linamarin and lotaustralin from leaves of T. repens involved continuous extraction of an aqueous syrup with ethyl acetate. Later attempts to repeat this method failed to give satisfactory crystallization and it was replaced by two modified techniques in which the aqueous syrup was either extracted with alcohol-acetone or exhaustively boiled with ethyl acetate. Linamarin has also been extracted from tubers of Manihot esculenta by chromatography of an ethanolic extract with chloroform-methanol on silica gel and crystallization from ethyl acetate, and by an adaptation of the dry column technique of Loev and Snader. A method for the isolation of dhurrin from Sorghum vulgare involving chromatography on cellulose powder with butan-1-ol-water (9:1) has been described by Mao et al. and has now been adapted for the isolation of a mixture of linamarin and lotaustralin from leaves of T. repens. The rapid inactivation of glycosidases is an important preliminary to the extraction of glycosides from plant tissue. This problem has been avoided in the present study by the collection of leaves from a single clone, genotype AcAclili, which contained both linamarin and lotaustralin, but no detectable linamarase activity. Clarified, deionized extracts of these leaves were used in various unsuccessful attempts to isolate the cyanoglucosides by dry column chromatography and by extraction of dried residues with ethyl acetate. However, chromatography on columns of cellulose powder gave reproducible separations of the cyanoglucosides, which readily crystallized from butan-1-ol. Paper chromatography of the isolated mixture and densitometric scans of spots located with ammoniacal AgNO3 suggested that the mixture was approximately 56% lotaustralin, 44% linamarin. Linamarase activity of T. repens leaf extracts was assayed by hydrolysis of the cyanoglucosides and estimation of the evolved HCN. The 10-min reaction period, which was used, required a sensitive method for the determination of cyanide and this was provided by the modified pyrazolone method. This short hydrolysis period provided a better indication of initial velocity of the reaction than the periods of one hour or more, which have been used in other studies. The reaction velocity was proportional to linamarase concentration and the optimum pH for maximum linamarase activity was 5.0. The linamarase assay has been used in a study of the properties of the enzyme in white clover and in comparisons of plants containing different alleles of the Li locus. The results of these experiments will be published elsewhere.