This paper identifies neighbourhood effects in consumption using the randomized nature of the Progresa programme. Recent studies establish that the programme affects the consumption of both eligible and neighbouring ineligible households but the underlying mechanism of the spillovers is not fully understood. I use disaggregated consumption data to distinguish between changes in consumption which result from changes in neighbourhood consumption and changes in consumption which are a result of income transfers between households. Using a flexible demand model that accounts for total expenditure, prices and household characteristics, I find that neighbourhood consumption has a substantial effect on household consumption choices