Knowing which restoration approach provides the best returns on investment for accumulating carbon is essential to foster restoration planning, financing, and implementation. We assessed the recovery of carbon stocks, implementation and land opportunity costs of forests established by natural regeneration and high-diversity native tree plantations. Our study was based on chronosequences (10-60 yr) of 12 naturally regenerating forests, 13 restoration plantations, and 5 reference forests located in Brazil’s Atlantic Forest. Restoration plantations accumulated approximately 50% more above-ground carbon than regenerating forests throughout the chronosequence. When controlling for soil clay content, soil carbon stocks were higher in reference than in restored forests, but they were comparable between plantations and regenerating forests. After 60 years of stand development, recovery of total carbon stocks in both restoration management types reached only half of the average stocks of reference forests. Total cost-effectiveness for carbon accumulation, including both implementation and land opportunity costs, was on average 60% higher for regenerating forests than for plantations (15.1 kgC.US$-1 and 9.4 kgC.US$-1, respectively). Both restoration management types had cost-effectiveness for carbon accumulation markedly lower than the price of carbon credits considered, so some voluntary forest carbon markets are not adequately priced to support restoration derived offsets. Although tree plantations initially had higher rates of carbon storage than regenerating forests, their higher implementation and land opportunity costs make them less cost-effective for carbon farming. Our results further suggest that carbon markets alone have a limited potential to up-scale restoration efforts in Brazil’s Atlantic Forest.