Abstract: The Daqing Mountains, a crucial natural ecological barrier in China’s northern frontier, is situated in the central section of the Yinshan Mountains in Inner Mongolia. As the most extensive, best-preserved and fastest-growing forest community in this region, the Pinus tabuliformis plantation forest plays a pivotal role in advancing carbon exchange and its affecting factors research in semi-arid regions. The net carbon exchange characteristics and their relationship with environmental factors during the growing season of P. tabuliformis (from May to October) in 2022 were analyzed on the basis of the carbon flux data obtained by using the eddy covariance method and meteorological data (T_a, P, PAR, T_s, VSWC, et al.) from the Hohhot National Climate Observatory. The results revealed that the diurnal variations of net ecosystem exchange (NEE) showed a U-shaped curve, with a carbon sink during daytime and a carbon source at night. The P. tabuliformis plantation ecosystem acted as a carbon sink from May to October, which varied seasonally with net ecosystem productivity (NEP)of 524.2 g C·m⁻²,ecosystem respiration (RE) of 611.8 g C·m⁻², gross ecosystem productivity (GEP) of 1 136.0 g C·m⁻². In August, higher precipitation significantly enhanced the daytime photosynthetic rate and respiration intensity of the P. tabuliformis plantation forest, however, the strongest carbon uptake occurred in September due to the sharp drop of respiration intensity in the daytime. At the half-hour scale, daytime NEE was primarily controlled by photosynthetically active radiation (PAR), with their relationship described by a rectangular hyperbolic model. The vapor pressure deficit (VPD) for photosynthesis during the daylight hours ranged between 10–20 hPa, while deviations from this range inhibited carbon uptake. Nighttime ecosystem respiration (RE) was predominantly affected by the soil volumetric water content (VSWC) at 5 cm depth. The relationship between RE and T_s at 5 cm depth followed an exponential equation (R² = 0.84), with a temperature sensitivity coefficient (Q₁₀) of 1.84. These findings not only deepen theoretical understanding of carbon cycling mechanism in semi-arid plantation forest ecosystems, but also establish a scientific foundation for sustainable plantation management practices, regional carbon flux modeling and global carbon budget assessment.