This study aims to select the global climate models GCMs-CMIP5(Global Climate Models-Coupled Model Intercomparison Project Phases 5) which is suitable in assessing of the impacts of climate change on irrigation water requirement (IWR) in the coastal area of the Mekong Delta. Statistical method has been used to evaluate the reliability of GCMs-CMIP5 model from to rainfall and temperature; and five GCMs were selected, including BBC-CSM1.1, GFDL-CM3, MIROC5, MRI-CGCM3, and NoESM1-M. The obtained data under emission scenarios (RCP2.6,RCP4.5 and RCP8.5) were then input into the CropWat model to assess the their impacts on IWR of rice and vegetables production to 2030s. The results have found that HadGEM2-ES proves the best capability of temperature simulation for the study area, with Pbias values of 0.21 0 C. IWR for rice and vegetables have the average values of 120.9mm, and 223.3mm, respectively in the period of 1990-2005. IWR in the future (2030s) under three scenarios are lower than that in the present.