Low temperature magnetic techniques provide useful tools to detect the presence of magnetite and pyrrhotite in sediments through identification of their low temperature transitions, to determine the amount of ultrafine-grained (superparamagnetic) material in sediments, and can potentially detect the presence of certain types of magnetotactic bacteria. Application of these types of experiments to nannofossil chalks from beneath the Barbados accretionary prism led to some unusual results, which are attributed to the presence of siderite. Thermal demagnetization of low-temperature remanence after cooling in zero field and in a 2.5 T field both displayed large remanence losses from 20 K to 40 K. Below 40 K, the magnetization of the chalks was much higher in the field-cooled experiments than in the zero-field-cooled experiments. Low temperature hysteresis experiments, made after cooling in a 2.5 T field, displayed offsets in magnetization parallel to the direction of the initial applied field, when measured below 40 K. The offset loops can be due to either an exchange anisotropy between siderite and magnetite phases in the sediments, a defect moment in the siderites, or a canted moment in the siderites. Apparent similarity between the low-temperature thermal demagnetization results from these siderite-bearing sediments, pure siderite, and pure rhodochrosite samples and the well-known 34 K transition in pyrrhotite should lead to caution in identification of pyrrhotite in marine sediments based on low-temperature remanence studies alone. ¿ American Geophysical Union 1996