To study variation in zinc efficiency (ZE) among current Chinese rice genotypes, a pot experiment was conducted with 15 aerobic and 8 lowland rice genotypes. Aerobic rice is currently bred by crossing lowland with upland rice genotypes, for growth in an aerobic cultivation system, which is saving water and producing high yields. A Zn deficient clay soil was used in our screening. Zn deficiency resulted in a marked decrease in shoot dry matter production of most genotypes after 28 days of growth. Genotypes were ranked according to their tolerance to Zn deficiency based on ZE, expressed as the ratio of shoot dry weight at Zn deficiency over that at adequate Zn supply. Substantial genotypic variation in ZE (50¿98%) was found among both lowland and aerobic genotypes. ZE correlated significantly (P <0.05) with Zn uptake (R 2 = 0.34), Zn translocation from root to shoot (R 2 = 0.19) and shoot Zn concentration (R 2 = 0.27). The correlation with seed Zn content was insignificant. In stepwise multiple regression analyses, variation in Zn uptake and Zn translocation explained 53% of variation in ZE. Variation in Zn uptake could be explained only for 32% by root surface area. These results indicate that Zn uptake may be an important determinant of ZE and that mechanisms other than root surface area are of major importance in determining Zn uptake by rice.