Measurements of temperature. the major species (N-2, O-2, CH4, CO2, H2O, CO, and H-2), OH, and NO are obtained in steady laminar opposed-flow partially premixed flames of methane and air, using the non-intrusive techniques of Raman scattering and laser-induced fluorescence. Flames having fuel-side equivalence ratios of phi = 3.17. 2.17. and 1.8 are stabilized on a porous cylindrical burner (Tsuji burner) in a low-velocity flow of air. Results are compared with calculations using a version of the Sandia laminar flame code that is formulated for the Tsuji geometry and includes an optically thin treatment of radiation. Because velocity profiles are not measured, the strain rate in each calculation is adjusted to match the measured profile of the mixture fraction. Measured profiles of temperature and species mass fractions are then compared with results of calculations using the GRI-Mech 2.11 and 3.0 chemical mechanisms, as well as a detailed mechanism from Miller. All three mechanisms give agreement with experimental results for the major species that is generally within experimental Uncertainty. With regard to NO formation. the relative performance of the three mechanisms depends on the fuel-Side equivalence ratio. GRI-Mech 2.11 gives reasonably good agreement with measured NO levels in lean and near-stoichiometric conditions. but it under predicts NO levels in fuel-rich conditions. GRI-Mech 3.0 significantly over predicts the peak NO levels the phi = 3.17 and 2.17 flames. but it yields relatively good agreement with measurements in the phi = 1.8 flame. The Miller mechanism gives good agreement with measured NO levels in the phi = 3.17 flame. but it progressively over predicts peak NO levels in the leaner flames, Comparisons of adiabatic and radiative calculations show that radiation can have a significant effect on the width and Structure of partially premix:ed flames. as well as on the levels of NO produced.