Virus is the most abundant biological component in marine and freshwater environments. Many studies have indicated the ecological importance of viruses in the primary production and microbial food web in marine environments. Repeated cycling of organic materials in the bacterial-viral loop causes the bacteria to be efficient sinks for C and regenerations of N and P in the environment where viruses are important agents of microbial mortality. However, precise enumeration of viruses in soil environment has been unsuccessful due to firm viral adsorption to soil particles. Floodwater in rice fields is an aquatic environment and the environment free from viral adsorption to soil particles. However, no study has been conducted on the viral ecology in the floodwater. As the floodwater ecosystem in rice fields is directly influenced by field management, this presentation enumerated the viral abundance in the floodwater of a Japanese rice field under a long-term fertilizer trial to evaluate the effects of the type of fertilization and the growth stage of rice plants. This is the first study on the viral ecology in the floodwater. The study field was located in the Aichi-ken Anjo Research and Extension Center, Central Japan (latitude 34.8° N, longitude 137.5° E). The field has been subjected to a long-term fertilizer trial since 1925. Four plots [a plot without fertilizers, a plot with chemical fertilizers, a plot with chemical fertilizers and lime, and a plot amended with chemical fertilizers and 22.5 tons ha^-1 compost (CM plot)] were chosen in this study (3.6Χ7.3 m² to 3.6Χ9.1 m²). Rice plants (Oryza sativa L. cv. Nihonbare) were cultivated in the respective plots under conventional management. Midseason drainage and the drainage for harvesting rice were performed in mid July and on September 28, respectively. Floodwater was collected at the midway of four sides of each plot along the footpaths once every week or two weeks during the cultivation period (11 times from June 15 to September 22). The sample water was passed through a 47 μm sieve, fixed with glutaraldehyde, and kept at -80 ēC until viral and bacterial enumerations. Viruses were first extracted with 3.6 % nutrient broth (NB) solution under ultrasonication for 2 to 3 min for desorbing viruses from suspended particles. Then, they were separated from suspended particles by centrifugation (8000 rpm for 10 min) and filtration with 0.45 μm Nuclepore filter successively. Viruses were stained with SYBR Green I for 15 min at 80 ēC, and enumerated with a fluorescent microscope. Separately, viruses in free form were determined after filtration with 0.45 μm Nuclepore filter without extraction with NB solution. Bacterial abundance in each floodwater sample was enumerated with SYBR Green I for 30 min at room temperature. In addition, composite fresh water samples from four sites were diluted sequentially and subjected to MPN method for enumerating bacteriophages against 18 bacterial strains that had been isolated from the floodwater of the rice field (6 α-, 2 β-, and 2 γ-Proteobacteria, 4 CFB members, a Bacillus sp. and 3 high GC Gram-positive bacteria). Bacterial abundance in the floodwater ranged in the order from 10⁵ to 10⁸ mL^-1, and there was no significant difference in bacterial abundance among the plots. It tended to increase with time after transplanting first until the beginning of July and was more numerous on June 28 and July 7 than on the other sampling time. Viral abundance in the floodwater was significantly correlated with the bacterial abundance, and ranged in the order from 10⁶ to 10⁹ mL^-1. Virus-to-bacterium ratios fell within the range from 0.11 to 72, and they tended to be larger during the middle growth stage of rice plant after midseason drainage when the viral abundance was smaller in the floodwater. Both viral and bacterial abundance also took a significant relations with the absorbance of water sample at 660 nm, indicating their adsorption to suspended particles as main way of presence in the floodwater. And viruses in free form (filterable with 0.45 μm Nuclepore filter) amounted to 10⁶ to 10⁷ mL^-1 without any seasonal variation. Fourteen to 73 % of floodwater samples contained more than a bacteriophage mL^-1 against the bacterial isolates (35 % on average). Among the tested host bacteria, bacteriophages infecting three strains of Sphingomonas sp., Enterobacter sp. and Cytophaga sp. often amount to over 10² to 10³ mL^-1 in the floodwater of the rice field. In general, bacteriophages tended to be abundant after midseason drainage.