The mode of action of triclopyr (3,5,6-trichloro-2-pyridinyloxyacetic acid), a broadleaf herbicide, is unknown. No data concerning the fate of triclopyr in plants have yet been published. This communication summarizes a comparison of the uptake and metabolism of triclopyr and 2,4-D (2,4dichlorophenoxyacetic acid) using cell suspension cultures of soybean, Glycine max, var. Harcor. Soybean cells were maintained in aseptic shake culture, on the following media: Gamborg’s B5 (Gamborg, 1975) supplemented with 2,4-D (2.3 PM), indole-3-acetic acid ( 2 . 9 p ~ ) , kinetin ( 4 . 6 ~ ~ ) and sucrose (2% w/v) at pH 6.5, or Miller’s medium (Miller, 1963) supplemented with naphthylI-acetic acid (1 1 PM), kinetin ( 2 . 6 ~ ~ ) and sucrose (2% w/v) at pH 5.8. Cells were subcultured (10ml into 40ml) at 14-16 day intervals. Seven days after subculturing, [2,6-’4CZ]triclopyr (0.82 pCi, 15.6 mCi/mmol) or 2,4-dichlorophenoxy[2-’4C]acetic acid (0.85 pCi, 13.4mCi/mmol, diluted from 55mCi/mmol, Amersham) was added in 50% aq. ethanol. After 6 h, 3 days, 7 days and 21 days, cultures were filtered under vacuum through nylon mesh. Cells were ground (liquid N,) and extracted (80% methanol, 3 x 15 ml), cell debris was sedimented by centrifugation and the supernatant solutions were taken. The I4C recovered from each culture (cells and medium) was quantified by liquid scintillation counting. Cell extracts and media were examined by t.1.c. Radioactive zones were located by radiography, cut from t.1.c. plates, suspended in scintillant and quantified by liquid scintillation counting. All incubations were performed in duplicate. In Miller’s medium: (i) 2,4-D was taken up slightly faster than triclopyr. (ii) Metabolites of both herbicides were largely retained in the cells, except the 2,4-D metabolites nos. 1 1 and 12 (see Fig. l), up to 30% of which appeared in the medium after 21 days. (iii) Metabolism of both herbicides continued until 21 days when no 2,4-D, and ca. 10% triclopyr, remained. (iv) Triclopyr was metabolized to two major compounds (Fig. l), having the following properties: (a) ion-exchange chromatographic behaviour and derivatization by diazomethane showed these metabolites to be acids; (b) acid hydrolysis (2 M-HCl, 80°C, 6 h) yielded triclopyr as sole product; (c) /-glucosidase treatment (citrate phosphate buffer, pH 4.5, 37”C, 18 h) left both metabolites unchanged; ( d ) the metabolites co-eluted on both silica gel t.1.c. and C,,-reverse phase h.p.1.c. with authentic triclopyraspartate (major) and triclopyr-glutamate (minor). Under the same conditions as above, 2,4-D was metabolized to several compounds (Fig. 1). By their chromatographic properties, the products of acid and P-glucosidase treatments, comparison with the triclopyr metabolites and standards described above, and precedents in the literature (Mumma & Hamilton, 1978), these metabolites were probably 2,4-D-glutamate (no. 8), 2,4-D-aspartate (no. 9), 4OH-2,5-D-O-glucoside (no. 10) and 4-OH-2,3-D-0glucoside (no. 11). The behaviour of cells grown on B5 differed significantly from that of cells grown on Miller’s medium in the following respects: (i) B5 cells rapidly released unchanged 2,4-D, but not triclopyr, after initial herbicide uptake. At 3 days and