To predict the effect of climate and CO2 on vegetation and soils, process-based models require accurate numerical representations of photosynthesis, respiration, turnover rates of below-ground carbon, and plant community change. We add 13C and 14C to the Lund-Potsdam-Jena (LPJ) dynamic global vegetation model (DGVM), which also includes representations of fire, land-use change, agriculture, and vegetation dynamics. Six sets of model simulations CO2 and 13CO2 fluxes are transported using the TM5 atmospheric transport model and evaluated by comparison to observations of CO2 and 13CO2 in measurements obtained from NOAA global monitoring sites. Updating the respiration and photosynthesis formulations in LPJ improves the seasonal cycle of simulated CO2 fluxes but does not improve simulated 13CO2 fluxes. Global annual land 13C photosynthetic fractionation and disequilibrium are similar to previous estimates by other DGVMs. To evaluate model performance for 14C, simulated soil and heterotrophic respiration 14C:C ratios are compared to observations from the International Soil Radiocarbon Database. Simulated soil Δ14C values are significantly higher than measurements as found in previous studies. Soil flux Δ14C values have less of a positive bias than soil Δ14C values indicating that, while soil residence times could be improved, some old soil carbon is very recalcitrant and may not need to be modeled to reproduce fluxes. This study emphasizes the power of using 13C and 14C to test whether model changes that improve bulk C are improving underlying processes such as stomatal conductance or soil turnover rates.