In this article I point to some common pitfalls and particular challenges in research on human trafficking. I start out by presenting some of the challenges in identifying observable populations and behaviours, arguing that primary data collection in the trafficking field should focus on former victims, and not current victims or persons at risk.

Thereafter I discuss some of the factors that have inhibited the development and use of explicit operational definitions of trafficking. Third, I present some of the challenges in identification of trafficking victims, when the victims themselves do not want to identify with the trafficking label. Finally, the usefulness of different research strategies in the trafficking fields for the current knowledge needs is discussed.