This paper describes a method to produce chemical reactor networks (CRNs) consisting of large numbers of perfectly stirred reactors (PSRs) from computational fluid dynamics (CFD) simulations to predict pollutant emissions from combustion systems accurately, flexibly, and efficiently using detailed kinetic schemes and the kinetic post-processor (KPP) developed at Politecnico di Milano. Benefits of the method described here include its applicability to a wide range of combustion systems, its ability to predict emissions of a variety of pollutant species, and its speed. CFD and CFD CRN simulation results of the Sandia D piloted methane air diffusion round-jet flame are successfully validated against experimental data for axial velocity, mixture fraction, temperature, and speciation, including CO and NO mass fractions. A CRN consisting of a large number of PSRs is found to be required to simulate the system accurately, while ensuring independence of the solution from CRN size. The results of CFD-CRN analysis for a 1114 PSR network are used to study the pathways (thermal, prompt, N2O, and NO2) by which NO and NO2, the constituents of NOx, are formed in the flame. Results of CFD CRN analysis show that NO is produced in the high-temperature (T > 1850 K) flame brush by a combination of the prompt, N2O and thermal pathways and in the intermediate-temperature (1000 0.43) region, where a low O atom concentration encourages a reversal of the prompt pathway (i.e., NO reburning), and in low-temperature (T < 1000 K) regions by the NO2 pathway, which oxidizes NO to NO2. Rate of production analysis, performed using CHEMKIN PRO at specified locations throughout the flame, shows that the trends of NO production and consumption observed in these simulations agree with expected and published results. Finally, the study predicts that, of the total NOx produced by the Sandia D flame, 47% is due to the prompt pathway, 32% is due to the N2O pathway, and 21% is due to the thermal pathway. As future steps in this work, the CFD-CRN method will be adapted and used to predict and study emissions from a range of more complex combustion systems.