S U M M A R YThe aim of this work is to calibrate a magnetic susceptibility field probe (SM30) in order to allow measurement on natural boulders of any shape and size. This calibration was performed through measurements on pebbles of different shapes, sizes and lithologies. We model a correction factor (geometric factor) that has to be applied to the SM30 measurement to obtain the real volume magnetic susceptibility of a sample of known volume. This geometric factor depends mostly on the volume of the sample, and to a lesser extent on its shape. We also present an original magnetic measurement scheme that provides, with three SM30 measurements at different distances, an estimation of the volume of a sample. By combining the two calibration models, it is possible to obtain the volume magnetic susceptibility of a natural sample with only three SM30 measurements, without requiring additional information such as sample volume. On the other hand, the calibration performed on a semi-infinite homogeneous body with variable measuring distance leads to a 2-D model of the SM30 response over its integration volume and allows the determination, with a sufficient number of discrete measurement at variable distance, of the thickness and susceptibility of the different layers of a composite body. Both types of calibration were successfully validated on natural or synthetic samples. Therefore the SM30 portable susceptometer appears to be a suitable instrument to perform in situ magnetic susceptibility measurement of individual boulders or to establish magnetic susceptibility profiles. In addition, the modest dimensions of the probe and the relative simplicity of the proposed measurement schemes should enable automation of the measurements that could find applications for robotic exploration of solid bodies of the solar system, for example. Volume magnetic susceptibility K is a dimensionless parameter defined as the derivative of the intensity of the magnetization induced in a sample (M = mV , m being the magnetic moment and V the sample volume) with respect to the applied magnetic field. This intrinsic property can be routinely measured in the laboratory with susceptometers, acknowledging the fact that it requires a separate volume measurement. Magnetic susceptibility measurements have numerous fields of application as they provide information about the concentration and nature of iron phases. Weathering profiles (Mathé et al. 1999), pollution monitoring (Heller et al. 1998), classification of granitoids (Ishihara 1977), correlation between sedimentary cores and palaeoclimatic proxy (Thouveny et al. 1994), ash-layer detection in deep-sea sediments core (Touchard & Rochette 2004), meteorite classification and aid in interpretation of magnetic anomalies are a few applications.In this paper, we present and calibrate a magnetic susceptibility measurement scheme that does not necessitate independent volume measurements and can be easily used in situ and automated. We show that this measurement scheme allows the determination of the ...