Pulsar timing: This method uses very precise timing of pulsar pulses to detect small bodies orbiting them. So far the only Earth-mass objects around another star have been found in this way. The method can thus find almost any planet, but the stars themselves are abnormal and so do not tell us about planets around stars like our own Sun.
Astrometry: Measuring precise positions of stars can detect planets if they are massive (like Jupiter) and distant from the star (like Jupiter), but this works only for stars that are close to us, and is especially suited to low-mass stars such as M dwarfs because the stellar motions are larger.
Infrared interferometry: This also works best for M dwarfs because in that case the contrast between the star and planet is less. This technique is most sensitive to distant massive planets that are bright in the infrared. NASA is aggressively pushing this method, with a mission called Planet Finder to be build in the next century, preceded by the Space Interferometry Mission (SIM), an astrometry satellite.
Micro-lensing: This novel method has the potential for detecting low-mass companions to stars that are very distant, especially if the planets themselves are large, but also, with luck, for Earth-sized planets. This technique's major drawback is that an observation cannot be repeated, so that statistical analyses must be relied upon. There are also potential ambiguities with the interpretation of an observation.
Ultra-narrow-band radio emission: This is just another way of saying "SETI,'' the Search for Extra-Terrestrial Intelligence. Obviously this method is a long shot, but if a signal were to be detected the information we could learn about the host planet are profound: orbital period, rotation period, eccentricity, obliquity, etc. G dwarfs are probably the best targets for this technique, for reasons explained in Soderblom & Latham (1993).
Radial velocity variations: This is the method now yielding the most detections. G dwarfs make good targets because they are reasonably bright and have lots of narrow lines. Close-in and massive planets are those most easily detected, and the current threshold for detection is about 10 m s-1.
The width of solar absorption lines is about 6 km s-1, and Marcy measures velocities to ~6 m s-1, or one part in one thousand. This has allowed detection of companion bodies smaller than Jupiter when they are close to the parent star. Several are found in distant, circular orbits, suggesting formation like that of our Solar System.
Photometric variations: This method looks for planets transiting a distant star. Obviously such events will be rare, at best, so to work this method must stare at a large number of stars constantly. The benefit, however, is considerable because this method can probably detect Earth-sized planets, and if it were to fail, that would still be significant because of the large sample surveyed. A NASA mission called Kepler has been designed to undertake just this study. It would survey about 5,000 solar-type stars for about five years. The achievable precision is about one part in 105, compared to a signal generated by an Earth-sized planet of about eight times that. It would provide a context into which to place our Solar System and would also yield an enormous database of high-quality photometry for solar-type stars.