Rice is of key importance to Indonesia's national and household level food security. The choice of tariff policy has important implications for consumers and producers with policy makers having to decide between the trade-offs implied for the various stakeholders. In this study we use a multi-market model to assess the impact of hypothetical rice tariff changes on household welfare and other variables of interest to rice policy-makers. A reduction in the rice tariff from 30 to 0% reduces rice supply and wheat demand and stimulates rice demand and soybean supply. Rice imports increase from 0.8 to 2 million tons. Rural households except for the Java-top income group, see incomes fall. In terms of purchasing power all households gain very significantly. Eliminating rice tariffs increases crop diversification and more so in those areas and for those income groups which started off least diversified. It is clear that the higher retail rice price resulting from a 30% ad-valorem tariff rate imposes significant cost on the 90% of Indonesian households, including most of the very poor households, who are net rice buyers. The implied income gains appear relatively modest but do accrue to middle and poorer households especially in Java. On the other hand an increase in the tariff from 30 to 50% eliminates rice imports, reduces soybean output and stimulates wheat demand. Rural households, except for the Java-top income group, see incomes rise although the effect is relatively modest. In terms of purchasing power households are all worse off.