An optical interferometer, with angular resolution of the order of a milliarcsecond and a collecting area of the order of twenty square meters, might well be constructed in space within the next twenty years or so. In addition to the many "conventional" astronomical problems that could be studied by such an instrument, interesting studies of other planetary systems could be made if certain design features were to be included. The principal requirements would be: (i) The ability to phase track on a reference object close by, (2) stable optical paths that would be constant over several hours to a fraction of a wavelength, (3) apodization of the entrance apertures to minimize sidelobes, (4) optical precision of 1/60 or better for scales a centimeter or larger, and (5) spectrophotometric capability. The stability requirements are challenging but appear to be technically feasible, the apodization should not be a problem, the optical quality is within the current state of the art, and the spectrophotometric capability is an automatic consequence of proper interferometric design. Such an instrument would detect Jupiter-like planets directly out to twenty parsecs or more at wavelengths 15000 or so, and earth-like planets could be detected out to ten parsecs distance, with reasonable integration times. Since the instrument would have spectrophotometric capability, once a planet had been detected and its ephemeris determined, one could integrate for considerably longer times to look for spectral features in the atmosphere. The A-band of oxygen and the Hartley bands of ozone are obvious candidates. The detection of these features would be most provocative, suggesting the presence of earthlike life, More generally, any spectroscopic evidence that a planetary atmosphere was far from chemical equilibrium would be evidence of more unfamiliar life forms. At the very least, one would be encouraged to make more intensive studies of such planetary systems, both in the infrared and at radio wavelengths. Conventional wisdom seems to have been that direct optical detection and study of planets was not feasible. This turns out not to be so largely because of the increased interferometric expertise that has developed in radio astronomy. Except for quantum effects that affect detectability, the methods of aperture synthesis carry over completely to the optical case.