The first tests of supereffective flame inhibitors blended with CO2 have been performed in methane-air laminar co-flow diffusion flames stabilized on a cup burner. The CO2 volume fraction required to extinguish the flames was determined for a range of added catalytic inhibitor volume fractions. When added at low volume fraction, the agents TMT, Fe(CO)(5), and MMT were effective at reducing the volume of CO2 required to extinguish the flames, with performance relative to CF3Br of 2, 4, and 8, respectively. This performance advantage of the metallic compounds is less than that determined in premixed or counterflow diffusion flames. Further, as the volume fraction of each metallic catalytic inhibitor was increased, the effectiveness diminished rapidly. The greatly reduced marginal effectiveness is believed to be caused by loss of active gas-phase species to condensed-phase particles. Laser-scattering measurements in flames with Fe(CO)(5)/CO2 blends detected particles both inside and outside (but not coincident with) the visible flame location for measurement points above the stabilization region. For Fe(CO)(5) addition to the air stream at 450 muL/L, the peak scattering cross section for vertically polarized light was 1660 times the value for room-temperature air. The first detailed numerical modeling studies were also performed for methane-air cup-burner flames with CO2 and Fe(CO)(5) added to the oxidizer stream and are used to interpret the experimental results. The role of particles was also illustrated by the numerical results, which showed that significant levels of supersaturation exist in the flame for several of the important iron-containing intermediates. This particle formation is favored in the lower temperature stabilization region of the cup-burner flames, as compared to the higher relevant temperatures of previously described counterflow diffusion flames. The results of this study indicate that the appropriate flame configuration for evaluating the effectiveness of some fire suppression agents must be carefully considered, since in those cases, different flame configurations can switch the relative performance of an agent by an order of magnitude. (C) 2004 The Combustion Institute. Published by Elsevier Inc. All rights reserved.