This work aimed at studying variations on the diversity and composition of the bacterial community of a rice paddy field floodwater, subjected to conventional management, namely by using the herbicide molinate. The promotion of the herbicide biodegradation either by the autochthonous microbiota or by a bioaugmentation process was also assessed. This study comprehended four sampling campaigns at key dates of the farming procedures (seeding, immediately and 6 days after application of the herbicide molinate, and after synthetic fertilization) and the subsequent physic-chemical and microbiological characterization (pH, DOC and molinate contents, total cells, cultivable bacteria and DGGE profiling) of the samples. Multivariate analysis of the DGGE profiles showed temporal variations in the bacterial community structure and the Shannon's index values indicated that the bacterial diversity reached its minimum at the molinate application day. The highest bacterial diversity coincided with the periods with undetectable concentrations of the herbicide, although microcosm assays suggested that other factors than molinate may have been responsible for the decrease of the bacterial diversity. The ability of autochthonous microorganisms to degrade molinate and the influence of the herbicide on the bacterial community composition were assessed in microcosm assays using floodwater collected at the same dates. Given molinate was not degraded by autochthonous microorganisms, and considering it represents an environmental contaminant, bioaugmentation microcosms were assayed aiming the assessment of the feasibility of a bioremediation process to clean contaminated floodwater. A molinate-mineralizing culture, previously isolated, promoted molinate removal, induced alterations in the autochthonous bacterial community structure and diversity, and was undetected after 7 days of incubation, suggesting the feasibility of the process.