This paper analyzes the experiences of demobilization and reintegration of about half a million ex-soldiers during the first part of the 1990s in Ethiopia. We use rural household data on ex-soldier and non-soldier households to assess whether reintegration has been successful. We find that the targeting efficiency of the demobilization program has not been as good as generally claimed. Ex-soldiers have welfare levels similar to non-soldiers. They are generally considerably better educated but have fewer assets than non-soldiers. Using a “treatment effects” model, we find that returns to labor and assets for ex-soldiers are indistinguisable from those of non-soldiers' families. This suggests that ex-soldiers have been successfully reintegrated in the rural economy, even though this means they are sharing the low standards of living of the rest of the rural population.