<p>This study describes a modeling framework, model evaluation, and source apportionment to understand the causes of Los Angeles (LA) air pollution. A few major updates are applied to the Community Multiscale Air Quality (CMAQ) Model with high spatial resolution (1 km <span class="wHYlTd z8gr9e">×</span> 1 km). The updates include dynamic traffic emissions based on real-time on-road information and recent emission factors and secondary organic aerosol (SOA) schemes to represent volatile chemical products (VCP). Meteorology is well-predicted compared to ground-based observations, and the emission rates from multiple sources (i.e., on-road, volatile chemical product, area, point, biogenic, and sea spray) are quantified. Evaluation of the CMAQ model shows that ozone is well-predicted despite inaccuracies in nitrogen oxide (NO<sub>x</sub>) predictions. Particle matter (PM) is underpredicted compared to concurrent measurements made with an aerosol mass spectrometer (AMS) in Pasadena. Inorganic aerosol is well-predicted while SOA is underpredicted. Modeled SOA consists of mostly organic nitrates and products from oxidation of alkane-like intermediate volatility organic compounds (IVOCs) and has missing components that behave like less-oxidized oxygenated organic aerosol (LO-OOA). Source apportionment demonstrates that the urban areas of the LA Basin and vicinity are NO<sub>x</sub>-saturated (VOC-sensitive) with the largest sensitivity of O<sub>3</sub> to changes in VOCs in the urban core. Differing oxidative capacities in different regions impact the nonlinear chemistry leading to PM and SOA formation, which is quantified in this study.</p>