Bus stops, as any other type of transport node, may be criminogenic places by nature. Evidence from North American and British studies has repeatedly shown that areas that ‘contain’ a bus stop are more criminogenic than those without (Levine and Wachs, 1986; Loukaitou-Sideris 1999; Loukaitou-Sideris et al., 2002; Newton and Bowers, 2007; Smith and Cornish 2006; Tsai et al., 2011). Yet despite such findings, the link between bus stops and crime has been controversial (Kooi, 2013) and highly dependent on North American and British evidence. If one randomly selects two areas, the first one containing a bus stop and the second not, is it more likely that the first area has more pickpocketing than the second one? Previous studies have attempted to isolate the effect of bus stops from that caused by the place’s attributes (where bus stops are located) on crime. Surprisingly, the nature of bus stops is often neglected in the analysis. This article addresses this issue by assessing the potential impact of passenger flow and vehicle frequency on the geography of pickpocketing.