Abstract The flow and transport of a non-reactive tracer and melt water was monitored in a heterogeneous coarse sandy unsaturated zone in southeastern Norway, during the snowmelt of 2001. Electrical resistivity tomography (ERT) as well as conventional suction cup techniques was employed. A frozen solution of NaBr in water was supplied as a line source on the ground surface above two parallel vertical profiles monitored by the two measurement systems prior to the onset of snowmelt. The two monitored vertical profiles were separated by approximately 1 m. The results were analysed by visual comparison of images and by the use of spatial moments analysis. The two measurement approaches showed that the system was affected by the presence of preferential flow paths during the early stages of the snowmelt, perhaps due to ice near the surface, but the major part of the plume moves uniformly later in the snow-melting period. After most of the tracer plume has reached the depth monitored by both systems (i.e. below 0.4 m depth) there is a good consistency between the two datasets. Spatial moment calculations on the basis of ERT cannot be used to describe the movement of tracer alone, as the resistivity is affected by changes in both saturation levels and tracer concentration. Nevertheless, ERT appears to be an appropriate method to characterise regions of localised high infiltration in this type of soil. The method therefore constitutes a possible alternative and supplement to suction cups in a monitoring system.