The contamination of groundwater by arsenic is currently a major concern in Bangladesh. Arsenic in groundwater was first detected in 1993 following reports of many people suffering from arsenical diseases. Further investigations showed the extent of the problem with large areas of the country's water supply being affected and millions of people at serious risk of arsenic poisoning. Technology for arsenic removal from water already exists. However, the socioeconomic conditions which prevail in Bangladesh, do not permit implementation of this type of technology on grounds of cost. The main objective of this study was to develop a low cost technique for the removal of arsenic from contaminated groundwater using the naturally occurring iron, which is another water quality constraint in Bangladesh. The approach was to form arsenic-iron complexes by coprecipitation and adsorption of arsenic on iron. It has been demonstrated that provided the iron levels are sufficiently high (say >_ 1.2 mg/1), simple shaking of a container and allowing the arsenic-iron complex to settle out for 3 days could reduce the concentration of arsenic from 0.10 mg/l to Bangladesh standard (0.05 mg/1). In experimental program, As(III) form of arsenic was used as this form is more likely to be present in groundwater. From laboratory studies, it was shown that the removal rate was largely controlled by the Fe/As ratio, pH and the As concentration. Arsenic removal increases with increasing Fe/As ratio and is favoured by increasing pH in the range of 5 to 8. Separation of the precipitates was achieved by settlement. Following prolonged settlement, it was found that arsenic removal could exceed the removal achieved by filtration through a 0.45 μm filter paper. The experiments demonstrated that about 77% arsenic removal could be achieved from water containing 0.2 mg/l As(III), 4.0 mg/1 Fe at pH 7.5 by manual flocculation (1 min manual mixing) and 3 days settlement. The use of ordinary charcoal, which is cheap and easily available, was investigated for removal of arsenic and was found to be ineffective. From maps of the known distributions of As, Fe and pH, it was evident that 63% of the area in Bangladesh complied with the Bangladesh standard for arsenic. By interpreting the maps and applying the potential removal by coprecipitation-adsorption and settlement technique, it was estimated that a further 8% of area would comply with the Bangladesh standard freeing an additional 7 million people from arsenic contamination.