This paper analyses retinal cartography in terms of its reflection of anatomic data and its relation to several forms of geographic methods of map-making. It shows that the distances between anatomic landmarks of the eye are reasonably similar to the relative distances on the retinal drawing chart currently used. Two forms of geographic cartography--azimuth equidistant and orthographic--are described and compared with retinal cartography. The retinal drawing chart currently used most closely approximates an azimuth equidistant projection, which suffers from circumferential distortion, a fact that retinal surgeons must keep in mind. It is therefore recommended that the chart be modified to have equally spaced concentric circles and clearer identification of the ora serrata; the present accurate marking of anatomic landmarks, such as the equator and the posterior border of the ciliary body, should be preserved.