Electrical measurements are used in various fields of geoscience and technology, e.g. gas/oil exploration or landslide-barrier monitoring. Although clays are amongst the most conducting geomaterials their electrical properties are not yet fully understood. For example, in a recent high-level-radioactive-waste repository large-scale test, a bentonite barrier was monitored geoelectrically. To facilitate interpretation of the results, the reasons for the observed differences in the electrical conductivity must be understood (e.g. changes in water content, temperature, salinity of pore water, etc.). To improve understanding of the electrical properties of clay minerals, in situ measurements must be combined with laboratory measurements. In situ measurements allow the characterization of the material in its natural state and laboratory measurements, for small sample amounts, allow the user to vary relevant parameters systematically such as water content, temperature, the salinity of the pore water, or even the cation population if swelling clay minerals are present. In situ measurements using different electrode distances, from m to cm range, proved that small-scale investigations are essential because of small-scale material heterogeneities. In the laboratory, all the relevant parameters mentioned above can be controlled more easily for small sample amounts. In the present study three different small-scale devices (SSM1–SSM3) were compared. The geometry factor, K, was determined both by calculation and by a calibration against solutions of different conductivity. Calculated and measured geometry factors were in good agreement. SSM1 and SSM2 – both with four pin-shaped electrodes – were found to be particularly applicable for in situ measurements. SSM2, with point contacts at the tips of the pins, was considered to be an improvement over SSM1 because the effects of both water content and temperature gradients (which are particularly relevant near the surface) were less pronounced using SSM2. SSM3, in which the contacts are placed at the bottom of a 4.5 mL trough, proved to be useful when systematically varying all of the parameters influencing the electrical properties in the laboratory.