Planetary transits of bright stars, V < 10, offer the best opportunity for detailed studies of extra-solar planets, such as are already being carried out for HD209458b. Since these stars are rare, they should be searched over the entire sky. In the limits of zero read-out time, zero sky noise, and perfect optics, the sensitivity of an all-sky survey is independent of telescope aperture: for fixed detector size and focal ratio, the greater light-gathering power of larger telescopes is exactly cancelled by their reduced field of view. Finite read-out times strongly favor smaller telescopes because exposures are longer so a smaller fraction of time is wasted on readout. However, if the aperture is too small, the sky noise in one pixel exceeds the stellar flux and the field of view becomes so large that optical distortions become unmanageable. We find that the optimal aperture is about 1". A one-year survey using such a 1" telescope could detect essentially all hot-jupiter transits of V < 10 stars observable from a given site.