The air quality and fire management communities are faced with increasingly difficult decisions regarding critical fire management activities, given the potential contribution of wildland fires to fine particulate matter (PM2.5). Unfortunately, in model frameworks used for air quality management, the ability to represent PM2.5 from biomass burning (BB) is severely limited. Particularly uncertain is the formation of secondary organic aerosol (SOA). This is due in large part to incomplete identification and quantification of compounds emitted from fires and uncertainties in mechanisms leading to SOA formation under ambient conditions. Thus there is great need for improved emissions inventories and validated smoke models that better capture emissions of intermediate and semi-volatile organic compounds (I/SVOCs) as well as SOA formation and aging as a function of fuel type and burn characteristics. We assembled a uniquely qualified team to: 1) provide improved emission factors (EFs) with an emphasis on those critical to understanding SOA; 2) develop a detailed model to accurately represent SOA in smoke plumes; and 3) use the detailed model as a tool, along with measurements, to implement and deliver an operational modeling framework with an improved ability to predict SOA formation, and thus PM2.5, from wildland fires. We achieved significant improvement in the characterization of gaseous organic compounds, including IVOCs (more abundant than SVOCs), relative to what is currently in the NEI and EPA SPECIATE emissions inventories. Data from four techniques were synthesized into a single EF database that includes over 500 gaseous non-methane organic compounds to provide a comprehensive picture of speciated, gaseous BB emissions. Of the total gaseous EF, 6-11% was associated with IVOCs. These atmospherically relevant compounds historically have been unresolved in BB smoke measurements and therefore are largely missing from emissions inventories. We highlighted some challenges in scaling these laboratory-based EFs to field conditions, particularly given the large diversity of potential SOA precursors and their dependence on fuel type, burn characteristics, and particle loadings. We identified and prioritized a subset of compounds for consideration in air quality models. The identified compounds were screened for published SOA yields; 55-77% of the reactive carbon was associated with compounds for which SOA yields are unknown. Thus we developed reaction mechanisms and SOA arameterizations, needed for air quality modeling, using 0-D box models. While limited data precluded complete model representation of all compounds, we demonstrated the sensitivity of SOA and PM2.5 predictions to newly identified SOA precursors and an increase in total organic carbon. CMAQv.5.0.2 and v.5.2 were used to predict SOA and PM2.5 concentrations for August 2013 and 2015. Model results were compared with measurements, including from the Biomass Burning Observation Project (BBOP). PM2.5 concentrations were generally underpredicted and relatively insensitive to changes in SOA using CMAQv.5.0.2. However, PM2.5 concentrations showed a greater sensitivity to and contribution from BB-derived SOA precursors using CMAQv. 5.2. Our BB emissions updates within CMAQv.5.2 generally resulted in a reduction in negative biases in PM2.5, and at some sites an overestimation of PM2.5. Our integrated measurement-model approach advanced the understanding of key sensitivities and uncertainties for predictions of the contribution of wildland fire emissions to SOA. Future work is needed to: 1) scale laboratory-based EFs of speciated gaseous organic compounds to field conditions (allowing for PM2.5 concentration-dependent partitioning); 2) further investigate the chemistry of key new precursors we identified (models and measurements); and 3) better constrain SOA model parameterizations using newly available data.