We compared published estimates of the net fine root production (FRP) in forest sites to litterfall and aboveground net primary production (ANPP) to test whether annual rates of fine root and aboveground production vary together at global scales. We also compared FRP estimates to theoretical upper limits as defined by our previously published relationship between total root allocation (TRA, carbon allocated to FRP plus live—root respiration) and litterfall. Estimates of the carbon content of FRP in the total data set ranged from 25 to 820 g°m—2°yr—1 and were not correlated with annual litterfall of ANPP. Different methods used for estimating fine root production, however, showed contrasting results. Estimates derived using either sums of seasonal changes in fine root biomass ("Sequential Core" method) or differences between annual maximum and minimum fine root biomass ("Maximum — Minimum" method) were not correlated with either litterfall of ANPP. Sequential Core estimates were often high relative to predicted TRA values, whereas Maximum — Minimum estimates were generally <50% of TRA. The small number of FRP estimates derived from root—free cores ("Ingrowth Core" method) were not correlated with measures of aboveground production but were all well below predicted TRA values. In contrast to results of other methods, FRP estimates derived using ecosystem N budgets ("N Budget" method) were positively correlated with both litterfall and ANPP. Comparing FRP estimates based on N budgets with previous results of forest soil C budgets suggested that annual fine root production increases with aboveground production and that °1/3 of TRA is used for production of fine roots.