Recent episodes of high and volatile prices for grains such as rice have raised concerns about their implications for hunger and poverty. We model price relationships between international rice prices and 221 domestic prices in 47 developing countries that import rice. We use a threshold vector error correction model that accounts for transaction costs of trade in spatial price transmission, and an improved regularised Bayesian threshold estimator for threshold models. Our results show that threshold values are higher after 2008 than before, which suggests that transaction costs in international rice trade have increased in recent years. Threshold values are highest for Latin American countries followed by African and Asian countries, and higher for retail than for wholesale prices. Since 2008, price transmission is slower in countries that responded to high and volatile prices with domestic market‐based interventions such as price controls and faster in countries that responded by lowering tariffs and by implementing production support measures.