Laser-induced fluorescence (LIF) of OH is used to measure spatially resolved temperature and concentration profiles in premixed laminar flames of H2 burning in mixtures of O2 and N2O at 7.2 Torr. Potential sources of error in such measurements are investigated: optical depth; the detector spectral bias, time delay, and sampling gate; and rotational level dependence of the quantum yield for the OH radical. We explicitly demonstrate differences between LIF intensity measurements and the actual OH concentration profiles caused by the temperature dependence of the rotational level populations across the flame front. By varying the proportion of O2 and N2O in stoichiometric flames, burnt gas temperatures between 1200 and 2300K are obtained. Quenching measurements in the burnst gases of these flames show that quenching by atomic hydrogen can be important. In the burnt gases of the H2/N2O flame, the quenching does not significantly depend on rotational level for diagnostics with 5% accuracy. However, in the chemically interesting and important flame front, there is a significant variation in the quantum yield with rotational level, and in the flame front of atmospheric pressure flames such quantum yield corrections are likely to be important.