The process of rice domestication has been studied for decades based on changing morphological characteristics in assemblages of both macroremains, such as charred seeds and spikelet bases, and microremains, such as phytoliths, esp. bulliform and double-peaked phytoliths. The applicability of these indicators in determining if a specific assemblage is wild or domesticated, however, is rarely discussed. To understand the significance of these indicators in the determination of domestication, we collected 38 archaeological samples from eight Neolithic sites, dating from 10-2ka BP, in the lower Yangtze River region to analyze and compare the changes of these different indicators over eight thousand years. The data demonstrate that the comprehensive analysis of multiple indicators may be the best method to study the process of rice domestication developed thus far. An assemblage of rice remains can be identified as domesticated forms if they meet the following criteria simultaneously: 1) the proportion of domesticated-type bulliform phytoliths is more than 73%; and 2) the proportion of domesticated-type rice spikelet bases is higher than 75%. Furthermore, we found that each indicator tends to change steadily and gradually over time, and each stabilized at a different time, suggesting that the characteristics of domesticated rice developed slowly and successively. Changes of multiple indicators during the period between 10,000–2,000 yr BP indicate that the process of rice domestication in the lower Yangtze River region lasted as long as ca. 6,000 years during the Neolithic, and can be divided into three stages with the turning points in the middle Hemudu-late Majiabang culture (6,500–5,800yr BP) and the late Liangzhu culture (4,600–4,300yr BP).