Tree-ring samples were collected from Larix principis-rupprechtii trees at three sites at different altitudes on Luya Mountain, in the Shanxi Province of North Central China. We used a "signal-free" method to alleviate potential biases due to climate medium-frequency disturbances, and generated three standard (STD) chronologies. The mean tree-ring width, first-order auto-correlation (ac1), and standard deviation (SD) of the three chronologies decreased as the altitude increased, and the mean sensitivity (MS), absent rings percentage, and series inter-correlation (r) increased with increasing altitude. The low-frequency and high-frequency chronologies were generated by applying low-pass (f < 0.1037) and high-pass (f > 0.1037) filters. Generally, the climate sensitivity and high-frequency variability of tree-ring chronologies increase with increasing altitude, while mean tree-ring widths and low-frequency variability were lower at high elevations. Low-frequency chronologies from middle and low altitudes exhibit less variation; while high-frequency chronologies from middle and high altitudes are similar to each other. The chronologies of high and low altitudes have few common variations. The correlation function and response function analyses were used to identify the key climate factor limiting the growth of trees at different altitudes. A significant negative correlation was observed between tree rings at low altitudes and the minimum temperature in May, while standard and low-frequency chronologies were significantly negatively correlated with soil temperatures in May and June. This indicates that trees at low altitudes experience drought stress due to the warm, dry climate. The STD chronology is most significantly positively correlated with the maximum May temperature, and significantly negatively correlated with the maximum April temperature at the middle altitude site, which has a relatively cold and wet climate. The relationship between chronology and soil temperature is not significant. High April temperatures intensified drought conditions and suppressed the expansion and absorption of water by cambium cells. Greater moisture and higher temperatures in May were more beneficial for tree growth than high moisture and temperatures in April. The relationships between chronologies (STD and low frequency) and climate factors at the high-altitude site with seasonally frozen soil are not significant, while the low-frequency chronology is positively correlated with April soil temperatures. Low soil temperatures could suppress the activity of cambium cells and delay the beginning of the growing season; thus, leading to the formation of narrow growth rings. Generally, the environmental conditions became increasingly moist and cold as altitude increased, and the factors limiting tree growth changed from growing season drought conditions to low soil temperatures at the beginning of the growing season. It appears that the current warming trend favors coniferous forests at high altitudes. However, drought stress on the mixed broadleaf-conifer forest at the low altitude site may be intensified; this would provide a reasonable advice for the forestry management in arranging the forest pattern in different altitudes.