In the Ganges-Meghna-Brahmaputra (GMB) delta of West Bengal, India, and neighboring Bangladesh ingestion of arsenic-contaminated groundwater has caused serious and widespread human health effect. In West Bengal about 55%, and 34% of 90,000 hand tubewells we had so far analyzed from affected districts by FI-HG-AAS showed arsenic values above 10 mug/l and 50 mug/l, respectively, while in Bangladesh the values were 73% and 59%, respectively, out of a total of 27,000 samples we had analyzed so far jointly with the Dhaka Community Hospital. Altogether over 100 million people are potentially at risk in these two countries. The source of arsenic is geologic. Examination of the sediments (it = 2235) from 112 boreholes showed that 85 samples contained arsenic 10-196 mg/kg and opaque particles separated from the parent sediments contained arsenic up to 2778 mg/kg. Mineralogical studies from several laboratories clearly show that both pyrite and iron oxide contain very high concentrations of arsenic. If oxidation of pyrite is the source of arsenic in groundwater then there are two questions to be answered: (a) whether arsenic-rich pyrite is sufficient to account for the mass of arsenic mobilized in groundwater; and (b) whether increased pyrite oxidation has led to increased arsenic and sulfate concentrations. Ore microscopic examination using reflected light indicates abundant highly reflective opaque particles, suggesting the presence of sulfide minerals. Electron microprobe analysis of the pyrite particles revealed arsenic contents ranging from 0.07 to 1.36 wt%,,. Gypsum may be a product of oxidation of sulfur in pyrite to sulfate followed by dissolution to calcite common to borehole samples studied. We have also identified framboidal and other secondary pyrite; this might partially explain the low sulfate concentrations in groundwater. Interviews of thousands of old people in affected area lead LIS to conclude that arsenical skin lesions were not present in those areas before the early eighties. Also, we have found that initially low-arsenic groundwater from some tubewells now contains unsafe arsenic concentrations. We believe that a single mechanism is not responsible for the release of arsenic from groundwater aquifer sediments in this region.