AbstractIn magnetic routine interpretation the comparison of two‐dimensional model curves with measured magnetic anomalies is widely used for an approximate evaluation of the position and depth of magnetic models.Before starting an interpretation of a survey by means of two‐dimensional models, it is very useful to have an idea of the shape of anomalies caused by extended but finite bodies, taking into account various strike directions: Three sets of anomalies of thin plates (horizontal length 19, downward length 9, width 1) dipping 30°, 60°, and 90° resp. for various strike directions and an inclination of 20° were computed. Some of these anomalies, e.g. those with nearly N‐S strike direction look rather complicated, and at the first glance one would not expect that they are caused by such simple bodies.Several profiles crossing the computed anomalies perpendicularly were interpreted two‐dimensionally. For less extended anomalies the depths determined for the top of the plates are 10‐20% too small, the magnetization amounts to 50–75 % of the value of the finite bodies. The interpretation of the profiles covering more extended anomalies gave very accurately the same values for the position, depth and magnetization for the two‐dimensional body as for the original three‐dimensional model.Anomalies of vertical prisms with varying extensions in they‐direction were computed. Their differences in amplitude and in the distance maximum‐minimum show that interpretation of short anomalies by two‐dimensional methods yields depth errors of up to 20 percent.To see the possibilities of the separation of superimposed anomalies dike anomalies were added to the anomaly of a broad body in great depth and several attempts were made to interpret parts of the composite anomalies. The interpreted bodies lie too deep. In complicated cases the depth values can have large errors, but experienced interpreters should be able to keep the errors in the range of one