Forests play a major role in the global carbon cycle. Understanding the dynamics of the forest carbon cycle and its driving factors is challenging. This study utilized dendrochronology and long-term (2003–2017) eddy covariance (EC) carbon flux data to investigate the relationships between tree growth and gross and net ecosystem productivities (GEPEC and NEPEC) in different-age (15-, 42- and 78-year old) pine plantation forests in the Great Lakes region in eastern North America. Tree growth in these different-age pine forests was significantly (p < 0.05) correlated with observed annual GEPEC values, while coherence between tree growth and NEPEC was relatively poor. Current-year and 1-year lagged ring-width chronologies and climate variables, including spring (April–May) temperature (TSPR) and Standardized Potential Evapotranspiration Index (SPEISUM) over the summer months (June–August) were used to test ten different linear regression models to simulate tree-ring-based GEP (GEPTR) values at all three sites. This analysis showed that current-year growth was the best predictor of GEPTR at all three sites, when compared to observed GEPEC, except during drought years, when GEPTR was underestimated. Current-year tree growth models were then used to reconstruct GEPTR over the life span of each stand. These reconstructions showed low GEPTR values from 1978 to 1988 and from 2002 to 2007. Low GEPTR in late 1970s occurred in response to below average temperatures when there were no major drought periods, while low GEPTR in early 2000s occurred following drought-like conditions in 2002. However, in recent years relatively higher GEPTR was observed at all three different-age forest sites. This interdisciplinary study will help to improve our understanding of carbon exchanges and the key environmental controls and associated uncertainties on tree growth in these different-age plantation stands in eastern North America. It will also help to determine how these forests may respond to climate change.