Because climate change impacts different ecosystems and geographies in myriad ways, it is important to study how forests in different regions respond to changing climatic conditions. In northwestern China, the transition in recent decades from a warm and dry climate to a warm and wet climate is well documented. However, the response of forests to this transition is poorly understood. To better understand these changes, we collected tree-ring samples of Larix sibirica from 12 plots on the southern slope of the central Altai Mountains and established site-specific tree-ring width residual chronologies. Through hierarchical cluster analysis, we could divide the sites into two levels: high-elevation level (HEL, >1900 m) and low-elevation level (LEL, <1900 m). We then used the first principal component of each level as the regional residual chronology and performed a moving correlation analysis with climate factors (monthly temperature, monthly precipitation, scPDSI) over the period 1957–2016. We found that changes in HEL trees are mainly related to temperature: from 1957–1967, growth was positively correlated with the temperature in July of the previous year. From the 1960s to present, growth was negatively correlated with the temperature in April of current year. Since the 1980s, however, temperature in current year January has been the most important factor limiting growth. The relationship between LEL tree growth and climate has been stable throughout the period of observation, and is significantly and positively correlated with moisture. We predict that the radial growth of HEL trees will be inhibited by continued warming, whereas LEL trees will benefit from improved moisture conditions in the future. Our findings help to provide a more comprehensive understanding of the effects of climate change on forest ecosystems, and provide a scientific reference for forest management and carbon sequestration and emission mitigation policies.