3 paddy soils developed from typical parent material in Hunan province were used as the tested soil in this study, such as the reddish clayey soil, granitic sandy soil and alluvial sandy soil respectively from the quaternary red soil, the weathered granite and the river alluvium. Pot experiment was conducted to investigate the effects of arsenic on the growth, yield and uptake of arsenic in early rice by treating arsenic (0, 5, 10, 20, 30 mg/kg).The results showed that with treatment of 0–10 mg/kg, no significant effect on the indexes of rice growth and development was observed. While with arsenic treatment between 20–30 mg/kg, the tiller number and rice yield were significantly lower than that of the control: no harvest from alluvial sandy soil, 47.2%–84.8% lower for granitic sandy soil, and 24.8%–48.6% for reddish clayey soil was respectively. The arsenic content in rice straw and milled rice increased with the treatment arsenic concentration. As the addition arsenic above 20 mg/kg, the arsenic content of straw and milled rice all increased more than 100% when compared with the control treatment for the 3 paddy soil, of which the most amplification was alluvial sandy mud. Correlation analysis showed that the plant height, tiller number and dry weight of rice were correlated with soil available arsenic content in a negative way. Arsenic toxicity in alluvial sandy soil was the most serious among 3 paddy soils.