Long-term GPS data analysis was performed in order to analyze the seasonal variation of tropospheric delay and the success rate of the ambiguity resolution. For this analysis, a total of 57 stations including 10 IGS stations in East Asia were processed together with double-differenced observables using Bernese GPS Software V5.0. The time span for this study ranges from 2002.0 to 2012.5 (10.5 years). The average baseline length is 339.0 km and the maximum reaches up to 2,000 km. The analysis is focused on two things: the annual variation of the tropospheric delay and the ambiguity resolution rate. The tropospheric delay is closely related to the weather condition, especially relative humidity, therefore it was estimated that the maximum would be in summer, while reaching its minimum in winter with the apparent seasonal variations. On the contrary, however, the success rate of the ambiguity resolution shows the opposite pattern: its maximum was in winter and minimum in summer. The fact seems to be induced by the surrounding conditions; that is, the trees thick with leaves near the GPS antenna interfere with GPS signals in summer. This seems to confirm partly that there is a distinct trend in the decreasing success rate since 2006 because the trees are growing every year. It is necessary to eliminate the factors that degrade the GPS quality and the tropospheric modeling for Korea needs to be studied further.