This paper analyzes the effect of uncertainty in several key parameters on the marginal costs of carbon sequestration in forests. These parameters include the land supply elasticity, which governs the conversion of land from agriculture to forests and vice versa; parameters of the forest biomass yield function; parameters of the forest carbon density function; and parameters of the costs functions for accessing inaccessible land. Monte Carlo techniques are thus used to turn the global forest model with no probability (e.g., Sohngen & Mendelsohn, 2003; 2007) into a proper probability model through Latin hypercube sampling. For this paper, we have restricted our analysis to consideration of probability distributions for only two of the parameters described above. Specifically, these are the parameters of the forest biomass yield function and the land supply elasticity. The importance index and the least square linearization are used to determine the relative contribution of input parameters to the model results. Five hundred model runs in one simulation were performed with covariability among the parameters. The Monte Carlo simulations indicated that most of the uncertainty in forest area in developed countries relates to uncertainty in parameters of the biomass function while in developing countries, where deforestation is more important (e.g., Brazil), the simulation showed the parameters of land supply elasticity to have the most important implications for carbon supply. These results are perhaps not too surprising but they do point to the need to empirically estimate land supply elasticities in regions like Brazil, where such estimates are not currently available in the literature. The results also provide information that can be used to estimate uncertainty intervals for carbon sequestration cost functions.