In the mid-to-high latitudes area, evidence for reduced sensitivity of tree growth to increasing temperature has been reported from multiple forests under the global warming background. This phenomenon, known as the “divergence problem” (DP), described the incapacity of temperature-sensitive tree-ring width to track increasing instrumental temperature measurements since the mid-20th century. In this study, the tree ring width at upper tree line was significantly positive relative to mean June temperature in the south slope of Altai Mountains. There is no report on the divergence response phenomena in this region. The time span of meteorological data from Katongkalagai of northeastern Kazakhstan is more than 70 years, much longer than that of western China, which benefited the testing of DP at the south slope of Altai Mountains. In autumn 2007, we took tree-ring samples of Larix sibirica Ledeb. with increment borer at the south slope of Altai Mountains, north eastern Kazakhstan. The tree-ring width series were detrended with negative exponent function and three tree ring width chronologies were developed by ASTAN program. Based on the meteorological data from Katongkalagai weather station, we analyzed the varying trends of precipitation and temperature since 1932. The annual precipitation of Katongkalagai has decreased by the rate of 8.3 mm (10 a)-1, but the decreasing trend could not reach the 95% confidence level. However, the annual mean temperature of Katongkalagai has increased significantly since 1932 by the rate of 0.208 ℃?(10a)-1, and winter temperature has increased by the fastest rate of 0.348 ℃?(10a)-1). This phenomenon, ‘warming in winter’, coincided to the uppermost contributor of global warming. Correlation analysis on tree-ring width and monthly precipitation and temperature of Katongkalagai from prior September to current September showed that tree-ring width was positive relative to mean temperature in June significantly with the coefficient over 0.60. The relationship was significative in plant physiology. Such relationship between ring width and temperature was also observed at the north slope and upper tree line at the south slope of Altai Mountains. The running correlation analysis between 31 a mean temperature in June and standardized chronology of tree-ring width showed that, with the temperature uprising, the correlation coefficient decreased slightly. The highest correlation coefficient is at the lower temperature stage before 1960s. In general, tree-ring width responded stably to the mean temperature in June. There seems no obvious “divergence response” in the south slope of Altai Mountains. The summer temperature was increasing slowly, in which the mean temperature in June at sampling site increased slightly, not enough to switch the growth response of Larix sibirica Ledeb.. However, the response of tree-ring width to temperature would change under the continued warming and drying pressures. The year of 1992 is an abnormal response year, that whether it is relative to Pinatubo volcano eruption is unknown.