Tobacco leaves were exposed to (14)C-labeled ethylene (3.7 x 10(-2) microliters per liter) in the presence and absence of unlabeled ethylene and other compounds. Most of the [(14)C]ethylene appears to be bound to displaceable sites. Lineweaver-Burk plots for a one-half maximum response in a tobacco leaf respiration test gave a value of 0.3 microliter per liter for ethylene, 50 microliters per liter for propylene, and 266 microliters per liter for carbon monoxide. Scatchard plots for displacement of [(14)C]ethylene from the site gave 0.27 microliters per liter for ethylene, 42 microliters per liter for propylene, and 746 microliters per liter for carbon monoxide. At 2%, CO(2) displaces about 35% of the bound ethylene, but increasing the concentration to 10% does not displace the remaining [(14)C]ethylene. A value of 3.5 nanomolar was calculated for the concentration of ethylene-binding sites available to exogenous ethylene. This does not account for the sites occupied by endogenous ethylene, and the total number of binding sites is probably somewhat higher. Using tissue culture material, the system was shown to be stable to freezing and thawing; and the pi-acceptors, carbon monoxide, cyanide, n-butyl isocyanide, phosphorous trifluoride, and tetrafluoroethylene, were shown to compete with ethylene for binding.