Smoldering combustion in wildland fires is a critical phenomenon that needs in-depth study because it can initiate with weaker ignition sources, can persist for long periods, is difficult to suppress, and can transit to flaming combustion. Cellulose, hemicellulose, and lignin are the major organic constituents within biomass, in varying proportions based on the type of fuel. The goal of this study is to computationally model and analyze smoldering combustion in mixtures of these constituents that represent real forest fuels. Smoldering combustion is modeled using a kinetic scheme that involves kinetic parameters and thermophysical properties. These data can vary depending on the source and the experimental setup used for measurements. Variations in input parameters significantly impact the output calculations. An uncertainty analysis estimated uncertainties associated with these inputs while a sensitivity analysis identified which parameters have higher influence on outputs. Kinetic properties are highly sensitive to the behavior of smoldering combustion and uncertainties associated with them are lower. In contrast, physical properties have higher uncertainties associated and are less sensitive to smoldering behavior. Further, fuel composition, density, oxygen concentration, and moisture content affect smoldering behavior, including propagation speed and peak temperature. Increases in lignin content decrease the propagation speed, while increasing hemicellulose content raises the propagation speed due to variations in rates of pyrolysis. Peak temperature rises with both increasing lignin and hemicellulose content, caused by the formation of ash. When the density of a mixture increases, propagation speed decreases and peak temperature rises. Accurately modeling smoldering in a given fuel requires characterizing whether moisture content causes expansion. In the context of smoldering ignition, radiative heat sources and hot metal particles can ignite a fire. Chemical kinetics drive the ignition via radiation, while physical properties drive ignition from conduction. Thus, fuels with varying physical and chemical characteristics will behave differently under different ignition sources. The insights gained from the model and these studies act as a novel framework for predicting smoldering combustion of forest fuels in wildland fires.