The effectiveness of arsenic removal plants (ARPs) to provide safe water was evaluated based on a study of 577 ARPs out of 1900 installed in 5 arsenic-affected districts of West Bengal, India. Out of 577, 145 (25.1%) were found in defunct condition. Both raw and filtered water from 305 ARPs were analyzed for total arsenic concentration. Forty-eight ARPs were installed despite raw water arsenic concentrations below the Indian standard (50 μg/L) and in 22 cases even below the WHO guideline value (10 μg/L). Among the 264 ARPs having raw water arsenic above 50 μg/L, 140 (53.1%) and 73 (27.7%) failed to remove arsenic below the WHO guideline value and Indian standard, respectively. The highest arsenic concentration in treated water was 705 μg/L. Analysis of 217 treated water samples for iron showed that 175 (80.6%) failed to remove iron below 300 μg/L. The treated water became coloured on standing 6 to 8 h, for 191 (44.2%) ARPs and 25 (5.8%) produced bad-odoured water. Overall, the study showed that 475 (82.3%) of the ARPs were not useful. The reasons for ineffectiveness and poor performance of these ARPs include improper maintenance, sand gushing problems, a lack of user-friendliness and absence of community participation. A comparative study of ARPs in two different blocks (Domkol in Murshidabad district and Swarupnagar in North 24 Parganas) showed that 39 (80%) and 38 (95%) ARPs, respectively, were not useful. Further study in Gram Panchayet Kolsur, Deganga block, North 24 Parganas, showed that 14 (87.5%) ARPs were not useful. Proper watershed management with active participation from the villagers is urgently required for successful mitigation. Copyright © 2006, CAWQ.