We use original 2005 household survey data from Fiji and Tonga to estimate the impact of migration and remittances on income distribution and measures of poverty. Measures of inequality and poverty based on observed, with?migration income and remittances are then compared with those based on two without?migration and remittances scenarios; one in which remittances are treated simplistically as an exogenous addition to household income, and another in which counterfactual household incomes are estimated, taking account of what the migrant members would have earned had they not migrated. The estimated effects of remittances on poverty alleviation were found to be stronger when the more rigorous, counterfactual income estimates are used, but their effect on income inequality was found to be ambiguous. The extent to which migration and remittances alleviate poverty was also found to be more substantial in Tonga, the country with a much longer migration history and higher remittance dependence. Copyright © 2008 John Wiley & Sons, Ltd.