Because of India's size, the age of its most recent available poverty survey, and the availability of a more recent nationally representative survey, the World Bank utilized a survey to survey imputation exercise as a key input to estimate poverty in 2015. At the $1.90 per day international poverty line, the imputation model predicts a 2014-15 headcount poverty rate of 10 percent in urban areas and 16.4 percent in rural areas, implying a poverty rate of 14.6 percent nationally. The implied poverty elasticity with respect to growth is within the range of past experience, and states with higher GDP growth saw greater predicted poverty reductions. In validation tests, the model's predictions perform comparably to the current adjustment method when predicting into 2011-12 but are far more accurate when predicting into 2004-05. Three alternative specifications give moderately higher estimates of poverty. The results suggest that survey to survey imputation, when feasible, is a preferable alternative to the current method used to adjust survey-based poverty estimates to later years. [2]