We present four years (2005–2008) of biometric (B) and eddy-covariance (EC) measurements of carbon (C) fluxes to constrain estimates of gross primary production (GPP), net primary production (NPP), ecosystem respiration (RE) and net ecosystem production (NEP) in an age-sequence (6-, 19-, 34-, and 69-years-old in 2008) of pine forests in southern Ontario, Canada. The contribution of individual NPP and respiration component fluxes varied considerably across the age-sequence, introducing different levels of uncertainty. Biometric and EC-based estimates both suggested that annual NPP, GPP, RE, and NEP were greatest at the 19-year-old site. Four-year mean values of NEP(B) and NEP(EC) were similar at the 6-year-old seedling (77 and 66 g C m−2 y−1) and the 69-year-old mature site (135 and 124 g C m−2 y−1), but differed considerably at the 19-year-old (439 and 736 g C m−2 y−1) and the 34-year-old sites (170 and 392 g C m−2 y−1). Both methods suggested similar patterns for inter-annual variability in GPP and NEP. Multi-year convergence of NEP(B) and NEP(EC) was not observed over the study period. Ecosystem C use efficiency was correlated to both forest NEP(EC) and NPP(B) suggesting that high productive forests (e.g. middle-age stands) were more efficient in sequestering C compared to low productive forests (e.g. seedling and mature stands). Similarly, negative and positive relationships of forest productivity with the total belowground C flux (TBCF) to GPP ratio and with the ratio of autotrophic to heterotrophic respiration (RA:RH), respectively, determined inter-annual and inter-site differences in C allocation. Integrating NEP across the age-sequence resulted in a total net C sequestration of 137 and 229 t C ha−1 over the initial 70 years as estimated by the biometric and EC method, respectively. Total ecosystem C sequestered in biomass at the 69-year-old site suggested an accumulation of 160 t C ha−1. These three estimates resulted in a mean C sequestration of 175 ± 48 t C ha−1. This study demonstrates that comparing estimates from independent methods is imperative to constrain C budgets and C dynamics in forest ecosystems.