1. A bioassay method based on the gibberellin-induced release of reducing sugars from endosperm, which was originally reported with barley by Nicholes and Paleg, was applied to rice endosperm and the results were compared with the rice seedling method, in which the elongation of leaf sheath was used. 2. Various compounds were screened for their possible effects on the release of reducing sugars from rice endosperms (Table 1). Organic acids (200, 500μg), amino acids (20, 200μg), vitamins (20, 200μg), auxins (2, 20μg) and kinins (2, 20μg) showed no positive effect in the rice endosperm test. Gibberellin A3 (GA3) and its degradation products, gibberellenic acid, and 2, 3, 7-trihydroxy-1-methyl-8-methylenegibb-4-ene-1, 10-dicarboxylic acid (dicarboxylic acid) stimulated the release of reducing sugars but allogibberic acid was inactive. Helminthosporol, which is a metabolic product of Helminthosporium sativum, and its derivative, helminthosporic acid, were active from the concentration of 10μg/ml. 3. Ten varieties of rice were tested for their responses to GA3. Their sensitivity is variable and may depend on their previous history as well as genotype. Aged seeds, whose germination percentage decreased, were less sensitive to GA3 than fresh ones (Table 2). 4. The isolated endosperms obtained from developing seeds of more than 18 days after flowering responded to GA3, but those from much more immature seeds did not (Fig. 3). Also in endosperms of germinating rice, GA3 was able to induce the release of reducing sugars (Fig. 4). 5. GA3 can be bioassayed in the range from 0.001 up to 0.1μg/ml using varieties Aichiasahi and Tanginbozu (Fig. 5). This eneosperm method is ten times more sensitive than the rice seedling test. 6. Extracts from the following plant species (Figs. 7-11) and human urine (Fig. 12) were examined for the presence of gibberellin-like substances: Pharbitis nil (immature seeds), Oryza sativa (shoots), Cryptomeria japonica (leaves), Equisetum arvense (sporophylls), Porphyra tenera, Chara coronata, and Saccharomyces sp. They were separated by paper chromatography and were found to contain active substances in terms of the release of reducing sugars from rice endosperms. However, the histogram pattern using the endosperm test was not always coincident with that using the rice seedling test. As a typical example, the histogram of gibberellin-like substances of extracts from rice shoots was shown in Fig. 9. By means of the endosperm test the remarkable activity appeared at two zones of Nos. 1 and 8, but in the seedling test the activity was observed at the zone of No. 4 with tailing in Nos. 2 and 3. 7. Fig. 12 shows evidence for the presence of gibberellin-like substances in human urine. With the endosperm test, an active zone appeared but by means of the seedling test weak active zones were obtained at different places on the histogram. 8. Attempts were made to compare GA3, its degradation products, gibberellenic acid, allogibberic acid, and dicarboxylic acid and helminthosporol on rice endosperm and seedling tests. The relative activities of these compounds to GA3 on the endosperm were different from those found for the seedling (Table 3). 9. These findings indicate that there are some differences in the response to gibberellins between the various tissues of the same species. 10. The rice endosperm test is very sensitive to GA3, but when the problem is concenrned with elongation, seedling tests are more suitable method than endosperm tests.