Until now little importance has been attached to psycho-environmental and even criminological research on pickpocketing (defined as a non-violent theft of objects from pockets or bags of an individual). Uzzell, Brown and Breakwell (2000) found that responsibility of being pick-pocketed is often attributed to the inefficient work of police rather than to an error of the pick-pocketed. However, the victimized person’s inattention is repeatedly implicated in cases of this theft. Pickpocketing is difficult to repress since the urban environment presents numerous factors facilitating pickpockets’ activity. One of these factors, the environmental overload, is likely to cause a fatigue and a distraction, making some individuals become easy prey for a pickpocket.Many findings in the field of the CPTED have shown that certain environmental characteristics can be favourable to the occurrence of specific crimes in an area. Criminals, acting in a rational way, research places to commit a crime according to their perception of opportunities and examining social and physical environment of places (Brantingham & Brantingham, 1995; Clarke, 1992). Nevertheless, common perception of crime occurrence risk rarely coincides with the reality of criminal events and is related to ecological labels (physical and social) rather than actual crime rates. Moreover, a potential victim’s risk perception is not determined by the reflection on opportunities that environment offers to a criminal (Chaguiboff, 1992). ObjectiveThis research aims to study pickpocketing by examining the interactions between the potential victim and the pickpocket; interactions that take place in the tissue of objective environmental settings. We attempt to establish the difference in perceiving the same environment by both actors: 1) by perpetrator, from the point of view of criminal opportunities that environment offers;2) by the potential victim from the point of view of pickpocketing risk that environment represents. MethodThe environmental characteristics of 23 sites situated in Paris and in Moscow, defined by high or low pick-pocketing rate, were submitted to direct observations. The observed characteristics included those identified in our interviews with 25 perpetrators as favourable for the pick-pocketing commitment. 459 passers-by in each site were interviewed on their in-spot pick-pocketing risk perceptions. In order to have the third point of view, interviews with 80 pickpocketing victims were conducted.ResultsWith regard to environmental characteristics of sites, the findings suggest that in both cities the pick-pocketing cold spots are defined by elements of discharged environment while the hot spots are defined by elements likely to cause environmental overload. Analyses provided evidence that passers-by’s risk estimations do not coincide with the actual pick-pocketing rates. The risk is estimated as “low” at sites presenting characteristics of social order and numerous signs of wealth. And vice versa, pick-pocketing risk is estimated as “high” at sites demonstrating social disorder symbols. However, the pickpockets view these environmental characteristics differently. They prefer to act in places presenting characteristics of social order since population feels there in confidence. On the contrary, they judge unsuitable sites demonstrating social disorder, as they generate avoidance behaviour. These findings question the Defensible Space Theory, previously challenged by Shaw and Gifford (1994).Almost 30% of explanations given by passers-by to their risk estimations concern the level of social density of site. Although the idea that “pickpockets act in crowds” is generally accepted, people don’t know why pickpockets prefer crowds and how they use this factor. Our interviews show that perpetrators are aware of and profit of this. The analyses of interviews with pickpocketing victims show that over 35% of victimisation scenarios included three elements: attention drawn to some elements of the environment, an environmental overload, a relaxing atmosphere of the place. Besides, analyses demonstrate clearly that pickpockets apply special diversions only in 30% of cases, using, in other cases, the situation and particularly their knowledge about how the environment influences the victim.The discussion of this and other results will focus on theoretical and practical outcomes for the research on crime perception and victimization. The added value of this study consists in understanding the complexity of an environmentally determined crime from several angles (the one of potential victims, the one of actual victims, and the one of perpetrators).