As part of a comprehensive, nationwide evaluation of geothermal resources for Japan, the first of the Curie point depth maps, covering the island of Kyushu, has been prepared. The map was created by inverting gridded, regional aeromagnetic data. Two satisfactory algorithms were developed to invert the gridded data based upon a distribution of point dipoles. The first algorithm estimates [Formula: see text],[Formula: see text], and [Formula: see text], the coordinates of the centroid of the distribution, by computing a least‐squares fit to the radial frequency of the Fourier transform; the second algorithm estimates centroid depth only by computing a least‐squares fit to the squared amplitude of the frequency estimates. The average depth to the top, [Formula: see text] of the collection of point dipoles, was estimated by a variation of the second algorithm. The depth to the bottom of the dipoles, inferred Curie point depth, is [Formula: see text]. The depth estimates are hand contoured to produce the final map. The Curie point depth map is then compared to regional geology and heat flow data, and to a limited set of gravity data. Good correlations are found between the Curie point depths and the heat flow and regional geology. A spatial correlation observed between gravity and Curie point depths is considered a secondary, structural effect. Locations of the currently operating geothermal power plants correspond to the shallowest Curie point depths. Based on these comparisons, we conclude that the methods provide geologically reasonable results which are usable in a nationwide geothermal assessment program.