The growing interest in urbanization problems is stimulating detailed studies of their effects on local climate change in the developed world. The absence of such studies in developing countries is restricting many decisions to be made and applied by policymakers. In one developing country, Turkey, results of the study of four urban stations and their neighboring rural sites for the 1951-1990 time period reveal that there is a shift towards the warmer side in the frequency distributions of daily minimum and 21.00 hr temperature difference series. This shift is an indication of urban heat island. The maximum urban heat island intensity trend that is obtained from the temperature differences database agrees well with Oke's (1973) formula for European cities. Seasonal analysis of individual 21.00 hr temperature series suggests that the regional warming is strongest in spring and weakest in autumn and winter. Urban warming is detected to be more or less equally distributed over the year with a slight increase in the autumn months. The Mann-Kendall trend test is applied to the temperature difference series, and the urban heat island effect is found to be significant in all urban sites. On the other hand, almost no significant urban effect on precipitation can be detected.