Arsenic contamination of the irrigation water-soil system is emerging as a big problem in Bangladesh threatening agricultural production, food quality and ultimately the health and well being of many millions of people. Elevated levels of As in soils and crops, especially rice, the staple food crop have been reported from different parts of the country. This has been, due, partly, to the use of high-As groundwater pumped out by shallow tube wells (STW). High As in rice grain can become a public health hazard. There is another potential risk, that of As affecting rice growth and yield, undesirable for Bangladesh with a high demand for rice. However, to date, little information on the growth retardation or yield loss of rice in farmers' fields in the As-affected areas of Bangladesh is available. We looked for indications of this in a high-As STW command area through a simple experiment in the wet season rice growing period, June to November, 2005. A 3-ha STW (32 m deep) command area in Paranpur of the As-affected district of Faridpur in central Bangladesh was the experimental location. Experimental plots, 4mx4m in size, in 10 fields across the command area with total soil As ranging from 15 to 57 mg/kg were set up. Transplanting with 25-30 days old rice seedlings(variety BR11) was done during 25-28 June. Soil samples at four different depths up to 60 cm at 15-cm intervals from the top were collected at the start of the experiment and analyzed for total, oxalate and dithionite extractable As, Fe and Mn. Arsenic was determined by the HG-AAS procedure and Fe and Mn in AAS. The collection of other data was scheduled to match the critical growth stages of the rice crop, i.e., mid-August (active tillering stage, AT), mid- September (panicle primordia initiation stage, PI), and mid-November (maturity). Soil pore-water was collected in mid-August when the soil was saturated with rainwater, and analyzed for As, Fe, Mn and P. Pore-water sampling could not be done later on as the soil started drying with the fast decreasing rainfall. At the AT and PI stages, agronomic data, such as, plant height, tiller no. per plant, root and shoot dry matter yields were taken. The crop was harvested on November 17, within a 5 sq m area per plot. Plant samples were taken for determining As in the roots, shoots, straw and grains. At the early stages of growth, the rice plants were very visibly poor in the high As plots. The pore-water As content ranged from 54 to 973 ug/L, which correlated well with the total soil As. Pore-water As appeared to be a good indicator of plant growth at the early stages, affecting all the plant growth parameters measured and also As uptake by the rice plants. However, as the soil started drying, we noticed the crop recovering, faster in fields on the higher positions of the command area landscape. We noticed a large, consistent decrease in the total soil As content with soil depth, indicating that As was not very downwardly mobile. The water regime, and by implication, soil reduction and associated mobility of As and its availability to the plants could be important factors regulating As uptake and its impact on plant growth. Soil As seemed to have no consistent effect on rice yield. However, although the crop recovered outwardly, maturity was greatly delayed in the high-As plots. About 40-50% of the grains in the high-As plots remained green, at the milk stage. Generally, grain yield was very poor, only 2.0-2.5 t/ha (rough rice), which was much less than about 4.0-5.0 t/ha that BR11 yields in a normal soil-water setting. This low yield was suspected to be due to the relatively high As in the soil across the command area. We hope to get a clearer answer to the question of whether As reduces rice yield through our planned study at the same site in the coming irrigated winter rice, when the soil will remain flooded with irrigation water throughout the growing season and the effect of uneven soil drying will be eliminated.