We investigate issues related to volatility and multi-generational gas-phase aging parameterizations affecting the formation and evolution of secondary organic aerosol (SOA) in models. We show that when assuming realistic values for the mass accommodation coefficient, experimentally observed SOA evaporation rates imply significantly lower “effective volatility” than those derived from SOA growth in smog chambers, pointing to the role of condensed phase processes and suggesting that models need to use different parameters to describe the formation and evolution of SOA. We develop a new, experimentally driven paradigm to represent SOA as a non-absorbing semi-solid with very low “effective volatility.” We modify both a box model and a 3D chemical transport model, to include simplified parameterizations capturing the first-order effects of gas-phase fragmentation reactions and investigate the implications of treating SOA as a non-volatile, non-absorbing semi-solid (NVSOA). Box model simulations predict SOA loadings decrease with increasing fragmentation, and similar SOA loadings are calculated in the traditional, semi-volatile (SVSOA) approach and with the new paradigm (NVSOA) before evaporation reduces loadings of SVSOA. Box-model-calculated O:C ratios increase with aging in both the SVSOA and the NVSOA paradigms. Consistent with box model results, 3D model simulations demonstrate that predicted SOA loadings decrease with the addition of fragmentation reactions. The NVSOA paradigm predicts higher SOA loadings compared to the SVSOA paradigm over nearly the entire 3D modeling domain, with larger differences close to the surface and in regions where higher dilution favors SVSOA evaporation.