With global warming upon us, it has be come increasingly important to identify the extent of this warming trend and in doing so be able to rank mean temperature changes in particular seasons and years. This requires a need for homogeneous climate data, which do not reflect individual anomalies in instruments, station locations or local environments (urbanization). Ac curate homogeneous long-term meteorological data helps show how temperature variations have truly occurred in the climate. Many possible factors contribute to artificial abrupt changes or sharp discontinuities in long time series data, such as the impact of station relocation, changes in observational schedules and instrumentation. Homogeneity adjustments of in situ climate data are very important processes for preparing observational data to be used in further analysis and research. Users require a well-documented history of stations to make appropriate homogeneity adjustments because precise historical back ground records of stations can provide researchers with knowledge of when artificial discontinuity has occurred and its causes. With out such de tailed historical data for each meteorological station, abrupt changes are difficult to interpret. Unfortunately, no homogeneity adjustments for temperature records have been con ducted previously in Tai wan, and present available sources of the history of Taiwan's meteorological stations exhibit in consistencies. In this study, information pertaining to station history, especially relocation records, is pro vided. This information is essential for anal y sis of continuous time series data for temperature and climate warming studies. Temperature data from several stations is given in this study to show how artificial discontinuity occurs due to station relocation. Al though there is no homogeneous adjusted climate data provided in this preliminary work, the summarizing of information regarding station relocations should be of assistance to future data users wanting to determine whether or not abrupt changes in climate data are artificial.