In the past 100 years, the rates of temperature changes have been obvious in the world. The objective of this study was to analyze the natural and artificial factors that could potentially influence annual mean temperature, annual mean maximum temperature (T□), annual mean minimum temperature (T□) as well as annual mean diurnal temperature range in Taiwan and to know the importance and the contribution of these factors. The product-moment correlation, the multi-regression model and the time series regression model were used to analyze the temperature data and potential factors affecting temperature changes for the period form 1961 to 2000. the results indicated that the artificial factors influencing the temperature changes were more important than the natural ones. Furthermore, the temperature change in big cities were significantly correlated to the amount of industries. Finally, the rare of increase or decrease for T□ and T□ differed among big cities with similar climate conditions and the correlations between temperature changes and the amount of different industries also differed. Above all, the economic development may affect the temperature changes in urban areas.