Dry etching of GaAs in plasma reactors is an important technology for the fabrication of optoelectronic devices. Currently, process optimization is done by trial and error, which is expensive in both time and money. Optimization techniques must therefore be developed based on a fundamental understanding of the related processes, and include a macroscopic understanding of the plasma chemistry. In this work, we examined the effect of gas flow rate on the peak intensity of the optical emission spectra of the chemical species in an etching chamber and on the etch rate of the GaAs substrate. From our results we developed a simplified model of the dry etching process of GaAs with Cl-2 in electron cyclotron resonance reactive ion beam etching that accounts for the degree of gas homogeneity in the reactor, especially the Cl-2 concentration. The effect of the reactor geometry and operating conditions on the performance of the dry etching process is well explained with this model, which can be used to optimize dry etching processes.