The quenching of N2(A3&Sgr;u+) by atomic oxygen in the auroral thermosphere offers an attractive method of determining atomic oxygen concentration, using the emisison rate ratio of the N2 Second Positive bands and the N2 Vegard-Kaplan bands. In the first applications of this method it was necessary to derive the quenching rate constant, the production rate ratio and the 0 profile from the measurement of emission rate ratio versus altitude. Even with more recent laboratory measurements of the quenching rates, different analyses of the same datasets gave rise to conflicting results. In this paper the consequences of including a requirement of diffusive equilibrium are considered; it is concluded that at times the 0 concentration at 130 km can be lower than that of CIRA 72 by a factor of 5, while at other times it can agree with the CIRA model.