Abstract: In order to study the internal water flow process of rice seed in presoaking and visualize the internal water transport process, an experiment of monitoring three varieties of rice seeds, namely SYAU No. 9816, Qi-shan-zhan, and Xiu-zi-nuo, during a 48h presoaking process with TD-NMR and MRI technology has been conducted. To learn the effect of the presoaking process on rice seed water distribution and water absorption, during the experiment period, the PDWIs of all samples have been obtained with SE pulse sequence, and the T2 spectral deconvolution images have been obtained with the CPMG pulse sequence every 6 hours. During the observation period, after presoaking for 0 hour, 6 hours, 12 hours, 18 hours, 24 hours, 30 hours, 36 hours, 48 hours, spin-echo (spin echo, SE) pulse sequence had been used to obtain proton density-weighted images of all the samples. CPMG (carr-purcell-meiboomgill sequence, CPMG) sequence had been used to obtain the transverse relaxation time T2 inversion spectrum of all the samples, so as to analyze the impact of seed soaking of the internal water distribution of rice seeds. NIUMAG MRI image processing software had been used on the 256 pixel × 256 pixel grayscale images acquired by the magnetic resonance imaging software for unified grayscale, pseudo color images, filtering and other processing, to adapt the images into forms that were more suitable for observation. The average value of the 3 CPMG pulse sequence value generated by the analysis software after repeated application of NMR signal at different corresponding time and the peak value had been calculated and the average value had been imported into NMR inversion software to obtain T2 spectral deconvolution. The experiment results showed that: MRI is an effective water detection technology, with whose help the internal water of rice seeds during presoaking can be monitored efficiently, accurately, without any loss. The use of T2 relaxation spectrum and proton density-weighted images helped to obtain a sample of internal hydrogen proton density and distribution, so as to reflect the moisture content and water flow information of the sample. PDWI can show the distribution of hydrogen proton within the samples. As in a PDWI, the part will be brighter if the density of hydrogen proton in that part is higher and the hydrogen protons within rice seeds mainly come from water, the brighter the part of the image is the higher level of water content in that part of the seeds it is. Therefore, with PDWI, the internal water distribution of rice seeds can be detected and explicitly shown. By observing the grayscale and the signal intensity curve of the three different kinds of rice after 48-hour presoaking time at various sampling points, the results found embryos had the highest moisture content, the seed coat came second, and endosperm had the lowest moisture content. A serial of pseudo-color maps of rice seeds obtained every 6 hours during presoaking can manifest the internal water dynamic flow process within the seeds. The map analysis showed that the water first penetrated the semi-permeable membrane of the embryos, went inside of the seed from the embryo end, penetrated the seed coat into the seed, and finally reached the endosperm. According to the T2 spectral deconvolution signal amplitude spectrum, water absorption of the rice seeds can be calculated. After each period of presoaking at various monitoring points, Xiu-zi-nuo showed the highest level of water absorption, Qi-shan-zhan came the next, and SYAU No. 9816 had the lowest level of water absorption. The paper analyzed the causes of the difference in water absorption: amylopectin has better water absorption than amylase, and Xiu-zi-nuo has the highest amylopectin content within the three different seed types, while SYAU No. 9816 has the highest amylose content. This conclusion was perfectly verified by the MRI results. The experiment results may provide valuable data for the module construction of water transport theory during rice seed presoaking.