What we know in 50 years depends a lot on what we can get off the ground in the next ten, IMO. I'll be (very) optimistic and assume that Kepler, SIM, and TPF (or their ESA equivalents) get off the ground in the next decade or so, and that it goes uphill from there.<br /><br />In 50 years we'll probably know of thousands--even tens of thousands--of planets in all size ranges. I think terrestrial planets will always be difficult to find and study, but we'll have plenty of those in pocket nonetheless. I wouldn't be surprised if we're able to pick out large moons around some exoplanet giants, either through timing or transit observations. Finally, I think we'll discover at least a few planets in the Large Magellanic Cloud; with super-telescopes of the future like the OWL and other interfermeters, I don't even think planets in M 31 or M 33 are out of the question at the tail end of the period.<br /><br />I also think that in 50 years, the following questions will be largely answered:<br /><br />1.) What is the metallicity "cut-off" for giant and terrestrial planet formation? Is there any at all?<br /><br />2.) Can stars significantly more massive than the Sun (say 2 solar masses and up) form planets? <br /><br />3.) Are planets as common around M-dwarfs as around sunlike stars?<br /><br />4.) How common are Earthlike planets? Mind you, by this definition, I mean planets that are roughly Earth-mass, in their planets habitable zone, and that have spectral signatures of oxygen, methane, or water vapor--singly or altogether.<br /><br />5.) Based on what we know about terrestrial exoplanets in the solar neighborhood, just how "special" is our own planet?