The endogenous ABA contents of dormant and nondormant barley grains were determined following application of different compounds to break dormancy. The chemicals used for breaking of dormancy in intact dormant grains were weak and strong acids, alcohols, hydrogen peroxide, cyanide, nitrate, salicylic acid, gibberellic acid and fusicoccin. The dormancy-breaking compounds could be classified into two major groups: compounds that caused a decrease in endogenous ABA (class I) and compounds which did not affect endogenous ABA (class II). Class I compounds included gibberellic acid, ethanol, hydrogen peroxide, nitrate, salicylic acid; class II compounds were fusicoccin, acid (H2SO4), sodium azide, n-caproic acid. In addition, these dormancy-breaking compounds were able to stimulate the germination rate when applied to embryos isolated from dormant grains. The concentrations necessary for stimulation of germination of isolated embryos were much lower than the concentrations for breaking the dormancy of intact grains. After embryos were isolated from dormant grains and incubated in water, ABA was determined in both embryos and in the incubation media. The class I compounds stated above also reduced ABA content in the incubation medium of isolated embryos, while class II compounds had no effect on ABA content of the medium. External application of ABA could overcome the effect of dormancy-breaking compounds of class I but not of class II. The results suggest that in the presence of the agents belonging to class II, ABA responsiveness of isolated embryos from dormant grains is decreased, compared to nontreated embryos.