An inversion algorithm has been developed to simultaneously infer the concentrations of two gases with overlapping spectral signatures using limb emission measurements. The algorithm is efficient and provides a solution in two or three iterations. It has been tested in a simulation study for the inference of stratospheric NO2–H2 and HNO3–CF(2CL2 Existing satellite instrumentation was the basis for errors used in the calculations. These included noise, scale and bias errors, angular registration errors, spacecraft motion effects, and effects due to a finite instrument field of view. It is estimated that concentrations of all four gases can be measured globally from a satellite with errors of less than 20%.