Journal of Ec. st. pp. 2-17Previous studies on food insecurity in many developing countries, including Ethiopia, have mainly focused on current food insecurity, lacking the ex-ante analysis. An under- standing of household vulnerability to food insecurity is critically important to inform the formulation of policies and strategies to enhance food security and reduce vulner- ability to food insecurity among smallholder farmers. This paper analyses vulnerability to food insecurity of farming households using the vulnerability as expected poverty (VEP) approach. The paper is based on cross-section data collected from a sample of 408 households in East Hararghe, Ethiopia, selected using a multi-stage sampling procedure. The factors which influenced vulnerability to food insecurity were analyzed using the Feasible General Least Squares regression model. The results of analysis indicate that vulnerability to food insecurity increased with the age of household head (P < 0.1), and family size (P < 0.01). It decreased with access to improved seeds (P < 0.01), adoption of soil and water conservation (P < 0.01), size of cultivated land (P < 0.1), and access to credits (P < 0.1). Based on the intensity of their vulnerability, households were grouped as chronic food insecure (24.27%), transient food insecure (11.77%), highly vul- nerable-food secure (18.38%), and low vulnerable-food secure (45.59%). Overall, about 54% of households were categorized as vulnerable to food insecurity. These included households who were food insecure at the time of the survey (36.02%) and those who were categorized as transient food secure group (18.38%). These findings imply that food insecurity policies and interventions in developing countries should focus not only on households that are currently food insecure, but also on those categorized as transient food insecure or households that are more likely to be food insecure in the near future.