There is no consensus among studies about how to define or measure poverty. Two widely used criteria to estimate poverty are the minimum expenditure criteria (income or monetary poverty) and the extent of deprivation regarding different dimensions representing quality of living of individuals (multidimensional poverty). This study compares income poverty and multidimensional poverty estimates across Indian states in recent years. Although both the poverty estimates indicate significant reduction in poverty, income and multidimensional poverty estimates differ significantly. Then the critical question is whether both measures identify the same set of households as poor or not.If alternate definitions of poverty identify different households as poor then we should be cautious about choosing the correct method to estimate poverty. In this context, we have studied the cases of two blocks, Salboni and Binpur I from Paschim Medinipur district in West Bengal. Using our own data from these two blocks, we find that poverty incidence calculated using two different measures for Salboni and Binpur I blocks give significantly different estimates. Even though most of the multidimensional poor households were also identified as income poor, not all income poor households were identified as multidimensionally poor.