Eddy covariance (EC) and biometric (based on tree diameter and tree ring measurements) methods are used extensively to estimate annual variability in forest productivity. Many previous studies obtained inconsistent findings between net ecosystem productivity (NEP) based on EC (NEP_EC) and tree biomass growth based on biometric methods, partly because of delayed responses in trees due to temporary non-structural carbohydrate (NSC) storage. Direct measurement of the amount of NSCs stored in trees is challenging; however, using carbon (C) isotope ratios in different tree components is considered an indirect and effective alternative approach for assessing NSC dynamics. We measured stable C isotope ratios in earlywood (EW) and latewood (LW) of rings of 70 trees (categorized as 4–10, 10–20, and > 20 cm diameter at breast height [DBH]) to test relationships between trunk-growth NEP_EC time series and annual tree biomass growth (based on tree ring width) in a young mixed forest in northern China over 13 years. Our results revealed a strong correlation (r = 0.691) between trunk-growth NEP_EC (March–August in a calendar year) and tree biomass growth. The change in correlation between current-year EW δ¹³C and LW δ¹³C of the previous year observed in trees with 10–20 cm DBH, which accounted for 51.8% of the total tree biomass growth, was consistent with the synchronicity of trunk-growth NEP_EC and tree biomass growth. Colder early growing seasons, which significantly increase the reliance of early-season structural growth on stored labile C accumulated in the previous year, may lead to postponed tree biomass growth responses related to NEP_EC. In conclusion, this approach, investigating C isotope analysis in EW and LW, helps elucidate the relationships between tree biomass growth and stand-level C fluxes in forest sites.