The Global Influenza Surveillance Network is crucial for monitoring epidemic risk in participating countries. However, at present, the network has notable gaps in the developing world, principally in Africa and Asia where laboratory capabilities are limited. Moreover, for the last few years, various influenza viruses have been continuously emerging in the resource-limited countries, making these surveillance gaps a more imminent challenge.We present a spatial-transmission model to estimate epidemic risks in the countries where only partial or even no surveillance data are available. Motivated by the observation that countries in the same influenza transmission zone divided by the World Health Organization had similar transmission patterns, we propose to estimate the influenza epidemic risk of an unmonitored country by incorporating the surveillance data reported by countries of the same transmission zone. Experiments show that the risk estimates are highly correlated with the actual influenza morbidity trends for African and Asian countries. The proposed method may provide the much-needed capability to detect, assess, and notify potential influenza epidemics to the developing world. The Global Influenza Surveillance Network is crucial for monitoring epidemic risk in participating countries. However, at present, the network has not can gaps in the developing world, principally in Africa and Asia where laboratory capabilities are limited. Moreover, for the last few years, various influenza viruses have been continuously emerging in the resource-limited countries, making these surveillance gaps a more immntial challenge. We present a spatial-transmission model to estimate epidemic risks in the countries where only partial or even no surveillance data are available. observation that countries in the same influenza transmission zone divided by the World Health Organization had similar transmission patterns, we propose to estimate the influenza epidemic risk of an unmonitored country by incorporating the surveillance data reported by countries of the same transmission zone. Experiments show that the risk estimates are highly correlated with the actual in fluenza morbidity trends for African and Asian countries. The proposed method may provide much-needed capability to detect, assess, and notify potential influenza epidemics to the developing world.