The magnetic signature of most kimberlite pipes is, at high magnetic latitudes, a circular anomaly. At lower magnetic latitudes, it becomes asymmetric; and at the magnetic equator, the anomaly is mostly negative. The shape of the anomaly is also influenced by the presence of remanent magnetization. For a vertical cylinder, the shape of the analytic signal of the magnetic field is nearly independent of field orientation and remanence and always results in a compact, almost circular anomaly. A simple pattern recognition technique, based on a first-order regression over a moving window, between the analytic signal of the observed magnetic field and the theoretical analytic signal of a magnetic vertical cylinder is an effective tool to identify potential targets. Results where the correlation coefficient between the analytic signal and the theoretical analytic signal within a moving window are above a certain threshold are retained, and additional criteria can later be used to refine the target selection. The method's practical utility is demonstrated by applying it to three different sites. The first is a well-documented area located at high magnetic latitude (Ontario, Canada), the second is at low magnetic latitude (West Africa), and the last one is an area where many pipes have a negative magnetization (Lac de Gras, Canada).