Despite the large individual benefits of guest work by the poor in rich countries, agencies charged with global poverty reduction do little to facilitate guest work. This may be because guest work is viewed as a repugnant transaction – one whose harmful side-effects might cause third parties to discourage it. This paper sets out six criteria for a transaction to be repugnant in consequentialist terms, and conducts uncommon tests for repugnance: It uses these criteria to formulate several empirical tests for the repugnance of guest work by Indian construction workers in the United Arab Emirates. It separates the effects of guest work from the correlates of guest work using a natural experiment that quasi-exogenously allocated guest work among a group of several thousand job applicants. The effects offer little evidence that guest work in this setting is typically the cause of repugnant consequences.