Although tourism obviously contributes to the economic growth of developing countries, the idea that it can be an efficient tool to alleviate poverty does not always receive unanimous support. The absence of a consensual methodology to measure the employee pro‐poor income impact of tourism has had a negative impact on two major levels: on research, to begin with, and on the acknowledgement of tourism as a tool to alleviate poverty. This paper discusses two recent tourism value chains, which allow the use and then the comparison of different poverty measures and then show that tourism impact strongly depends on the chosen poverty threshold. Copyright © 2013 John Wiley & Sons, Ltd.