Carbon dioxide (CO2) storage flux in the air space below measurement height of eddy covariance is very important to correctly evaluate net ecosystem exchange of CO2 (NEE) between forest ecosystem and atmosphere. This study analyzed the dynamic variation of CO2 storage flux and its effects on the carbon budget of a temperate broad-leaved Korean pine mixed forest at Changbai Mountains, based on the eddy covariance flux data and the vertical profile of CO2 concentration data. The CO2 storage flux in this forest ecosystem had typical diurnal variation, with the maximum variation appeared during the transition from stable atmospheric layer to unstable atmospheric layer. The CO2 storage flux calculated by the change in CO2 concentration throughout a vertical profile was not significantly different from that calculated by the change in CO2 concentration at the measurement height of eddy covariance. The NEE of this forest ecosystem was underestimated by 25% and 19% at night and at daytime, respectively, without calculating the CO2 storage flux at half-hour scale, and was underestimated by 10% and 25% at daily scale and annual scale, respectively. Without calculating the CO2 storage flux in this forest ecosystem, the parameters of Michaelis-Menten equation and Lloyd-Taylor equation were underestimated, and the ecosystem apparent quantum yield (alpha) and the ecosystem respiration rate (Rref) at the reference temperature were mostly affected. The gross primary productivity (GPP) and ecosystem respiration (Re) of this forest ecosystem were underestimated about 20% without calculating the CO2 storage flux at half-hour, daily scale, and annual scale.