This article: • Shows the variance of non-state actors in global crime governance and transnational governance in general, and shows that existent accounts fail to explain this variance. • Proposes a model of how we can understand the different roles of non-state actors, distinguishing normative from rationalist reasons for non-state actor involvement. • Compares different forms of current global crime governance (human trafficking, conflict diamonds, money laundering, cybercrime) to explore the validity of the model. • Shows that non-state activism and public debate are usually only related to a specific type of crime, turning a 'blind eye' to other forms of crime and their governance. • Argues that this creates problems with regard to oversight and discussion of global crime governance, exemplified with regard to intelligence surveillance via internet traffic. Global crime governance has become a major area of international activity, including a growing number of public and private regulatory efforts. Yet it is puzzling that a considerable variance exists in how state and non-state actors interact: non-state actors have been important agenda-setters in some issue areas, while they have been absent in others. Sometimes they are implementation bodies, sometimes they set regulations themselves. I argue that this variance is caused by issue characteristics: If an issue area is framed in a highly moralised way, it is likely that resulting nongovernmental activity can be explained by normative convictions, and in particular advocacy occurs frequently. If an issue area is framed in a technical way, resource exchange is central, and delegation to non-state actors becomes more prominent. A comparison of human trafficking, conflict diamonds, money laundering and cybercrime shows that this relation can be found on the global and national level.