Abstract Recent updates to U.S. regulatory guidelines for modeling impacts of single sources to secondary pollutants, such as ozone (O3) and fine particulate matter (PM2.5), require the use of established photochemical models with a complete treatment of gas-phase and aerosol chemistry as the preferred method for a Tier 2 demonstration approach. Such models include photochemical grid models (PGMs), as well as Lagrangian puff models with the requisite treatment of processes governing the formation of secondary pollutants from their precursors. SCICHEM is an open source reactive puff model that includes treatments for gas, aerosol, and aqueous-phase chemistry that are comparable to those in PGMs. The dispersion component of SCICHEM is based on the state-of-the-science puff model, SCIPUFF (Second-order Closure Integrated Puff). The model is a valuable tool that requires fewer resources and setup cost to run than PGMs to calculate the impacts of a single source or a small number of sources on downwind ozone and PM2.5. This paper describes a model inter-comparison study of secondary pollutant impacts due to emissions from a hypothetical coal-fired power plant stack. The models compared in this study include SCICHEM and the two most commonly used PGMs, CMAQ and CAMx. Stacks are located in 4 separate regions of the U.S. with distinct differences in chemical environments and geographical features. The key inputs (hypothetical stack parameters and emission rates, meteorology) to the three models are based on a common dataset for consistency. The source impacts calculated by the PGMs are based on zero-out simulations (i.e., with and without the source of interest) as well as the source apportionment tools available in CAMx, while SCICHEM directly predicts source impacts in a single simulation. Annual simulations are conducted to determine the frequency distribution of source impacts over the 4 regions. The comparison of model results shows both similarities and differences in the frequencies and magnitudes of peak impacts predicted by the three models. There is good agreement among all models in their predictions of 98th and 99th percentile impacts.