The subject of the invention is a method for geolocalization of one or more stationary targets from an aircraft by means of a passive optronic sensor. The sensor acquires at least one image I1 containing the target P from a position C1 of the aircraft and an image I2 containing the target P from a position C2 of the aircraft. The images I1 and I2 have an area of overlap. The overlap area has at least one target P identified which is common to the two images I1 and I2. The position of each target P is determined in each of the two images. The distance d is calculated between each target P and a point C, situated for example in the vicinity of C1 and C2, as a function of the angle β1 between a reference direction and the line of sight of the image I1, the angle β2 between the same reference direction and the line of sight of the image I2, of the position of each target P in the image I1 and in the image I2. The positions of the targets are calculated relative to the projection of the aircraft on the ground when at C, as a function of the distances d in a terrestrial reference frame.