Volatility determines the gas-particle partitioning of organic compounds. Volatility is thus a key property needed to understand the behavior of organic aerosol (OA) in the atmosphere. Various studies have been conducted to experimentally measure and numerically simulate distributions of OA volatility. The observed OA evaporation rates have generally been slower than the rates assuming instantaneous gas-particle equilibrium and volatility estimated from secondary organic aerosol (SOA) formation experiments. Particle-phase diffusion and/or low-volatility compounds, such as oligomers and highly oxygenated molecules, could limit the evaporation of OA, though the relative contributions of these factors are still uncertain. In this study, we conducted model simulations using a volatility basis set framework with the consideration of kinetic gas-particle partitioning, formation and dissociation of dimers, and particle-phase diffusion to reproduce observed evaporative behaviors of SOA formed from α-pinene ozonolysis and 1,3,5-trimethylbenzene (TMB)/NOx photooxidation. Based on simulations constrained by various volatility distributions derived from chemical analysis or heating experiments, we found that both dimerization and slow particle-phase diffusion contributed to the observed slow evaporation under dry conditions. In contrast, particle-phase diffusion did not practically inhibit SOA evaporation under humid conditions. The similarity of the fitted parameters, including dimer formation/dissociation rates and bulk diffusivity, for SOA from α-pinene and 1,3,5-TMB under dry conditions suggested that these processes are important for both monoterpene and aromatic SOA. Evaporation rates of SOA from α-pinene in this study were slower than the rates reported in previous experimental studies. This difference could be partly explained by differences in the experimental setups, including the treatment of organic vapors.