This study aims to provide a passive sampling approach which can be routinely used to investigate polychlorinated biphenyl (PCB) sources in rivers. The approach consists of deploying low density polyeth- ylene (LDPE) strips downstream and upstream of potential PCB sources as well as in their water dis- charges. Concentrations of indicator PCBs (iPCBs) absorbed in samplers (Cs) from upstream and downstream sites are compared with each other to reveal increases of PCB levels. Cs measured in water discharges are used to determine if released amounts of PCBs are compatible with increases revealed in the river. As water velocity can greatly vary along a river stretch and influences the uptake at each site in a different way, differences in velocity have to be taken into account to correctly interpret Cs. LDPE strips were exposed to velocities between 1.6 and 37 cm s-1 using a channel system built in the field. Relation- ships between velocity and Cs were established for each iPCB to determine the expected change in Cs due to velocity variations. For PCBs 28 and 52, this change does not exceed a factor 2 for velocity variations in the range from 1.6 to 100 cm s-1 (extrapolated data above 37 cm s-1). For PCBs 101, 138, 153 and 180, this change only exceeds a factor 2 in the case of large velocity variations. The approach was applied in the Swiss river Venoge to first conduct a primary investigation of potential PCB sources and then con- duct thorough investigations of two suspected sources.