AbstractRunoff generated from the Upper Indus Basin (UIB) mainly originates in the massifs of the Hindukush–Karakoram–Himalaya (HKH) region of Pakistan. Water supply in early spring depends upon the snow accumulation in the winter and the subsequent temperature. Seasonal temperature variations corroborate the contemporary dynamics of snow and glaciers. Recently, there has been increasing evidence of accelerated warming in high mountain areas, termed as elevation-dependent warming (EDW). We have identified trends, analyzed inconsistencies, and calculated changes in the maximum, minimum, mean and diurnal temperature range (Tmax,Tmin,Tmean, and DTR) at 20 weather stations during four-time series: 1961–2013 (first), 1971–2013 (second), 1981–2013 (third), and 1991–2013 (fourth). We employed the Mann–Kendall test to determine the existence of a trend and Sen’s method for the estimation of prevailing trends, whereas homogeneity analysis was applied before trend identification using three different tests. This study revealed that the largest and smallest magnitudes of trends appeared in the winter and summer, respectively, particularly during the fourth data series.Tmaxrevealed robust warming at ten stations, most remarkably at Gupis, Khunjrab, and Naltar at rates of 0.29, 0.36, and 0.43 °C/decade, respectively, during the fourth series. We observed thatTminexhibits a mixed pattern of warming and cooling during the second and third series, but cooling becomes stronger during the fourth series, exhibiting significant trends at twelve stations. Khunjrab and Naltar showed steady warming during the fourth series (spring), at rates of 0.26 and 0.13 °C/decade in terms ofTmean. The observed decreases in DTR appeared stronger in the fourth series during the summer. These findings tend to partially support the notion of EDW but validate the dominance of cooling spatially and temporally.