Accurate earthquake prediction is not yet feasible and some expects claim that it may be impossible to predict earthquakes. Therefore it has proven to be more beneficial to develop an infrastructure of seismic instrumentation used in conventional seismological investigations (e.g. crustal structure) but that will also function as an early warning system that can be used to reduce seismic hazard and mitigate the damage when large a earthquake occurs. The primary goal of an earthquake early warning system is to provide a warning, as quickly as possible (in real-time), that strong ground shaking will occur. This is accomplished by relying on the fact that P-waves arrive first and these waves typically have small amplitudes and cause minor, if any, damage compared to the large amplitude S-waves that arrive later because of lower velocity. Rapid analysis of the P-wave energy arriving at the closest stations to the earthquake’s epicenter can provide an estimate of the amount of shaking expected from the damaging S-waves; automatic tasks and preventive measures can then be taken before their arrival. The amount of time possible to issue a warning is clearly dependent on the time interval between the P-and S-wave arrivals that is directly related to distance from the earthquake source, i.e. the larger the distance, the greater the warning time. It is clear that small earthquakes pose less of a hazard than large events but even smaller magnitude earthquakes can cause significant damage when they are close. Therefore, two fundamental properties of the earthquake must be estimated as soon as the earthquake occurs: location and magnitude. Information regarding the feasibility of a real-time system in South Korea can be obtained from some results of the Japanese real-time system. Although Japan has the benefit of a very dense national seismic array, a large amount of instrumentation is not essential for a real-time system. Compared to earthquake magnitude, location is