The transit of a star by a dark companion or planet results in a characteristic colorimetric “signature,” which should be detectable by photometric means. This signature, which is due to differential limb darkening in red and blue light, takes the form of a slight shift towards the blue as the planet crosses the limb of the star, followed by an abnormal reddening during the transit of the center region of the star, and finally another shift towards the blue as the planet approaches the far limb. An analysis of this signature, the probabilities of detecting such transits (using the solar system as a model), and possible instrumentation for this purpose are discussed. A system designed to yield one or more planetary detections per year is believed to be feasible at moderate cost. Such a system would consist of three wide-field telescopes, at well-separated sites, slaved to a central computer. Each telescope would be equipped with a two-color photometer employing an image dissector under computer control for tracking and rapid switching between program stars without requiring frequent resetting of the instrument. Each telescope would survey an independent sample of stars on a rapid rotation cycle, until one of the telescopes detected an unusual transient, at which time all three would be used to provide additional data and increased reliability as a coincidence detection system.