An experiment was conducted at Nonaghata-Uttarpara village under Haringhata block in Nadia district during successive rabi seasons of 2007-08 and 2008-09 to study the arsenic contamination in vegetable pea cv. Karishma and to explore possibilities of offloading the accumulation through nutrient management using harvested surface water for irrigation. The experiment has been laid out with two irrigation sources in main plot I1: irrigation from Shallow Tube Well (STW) and I2: irrigation from Pond Water (PW) and four nutrient management options in sub-plot [N1: 100% RDF (60-80-90 kg ha-1 N, P2O5 and K2O), N2: 100% RDF with double the recommended dose of phosphate, N3: 75% RDF + FYM @ 10 t ha-1, N4: 75% RDF + FYM @ 10 t ha-1 with double dose of phosphate] in a split plot design, replicated three times. Results revealed that pod yield did not significantly vary with different irrigation sources, but arsenic accumulation and uptake was the lowest with irrigation from harvested rain water i.e. pond water. 75% RDF + 10 t ha-1 of FYM with increased phosphate recorded the maximum yield with the least arsenic accumulation (N4).