||Flare vendors recommend that cooling steam should always be supplied to the steam injection equipment on steam assisted flare tips. However, the nomenclature of "cooling steam" properly describes only one aspect of this steam flow. Not only does this steam provide thermal protection to the steam injection equipment in the high-heat zone near the tip exit, but this steam flow also ensures the steam supply piping remains heated, preventing condensate formation during large increases in steam flow rates. Unfortunately, during periods of low gas flow rates to the flare system (standby, purge-only, or small continuous flow rate modes) this cooling steam can reduce the hydrocarbon destruction efficiency of the flare system. Currently, the Texas Commission on Environmental Quality (TCEQ) is reviewing data from flare testing they commissioned through the University of Texas in 2010. Recent field observations by TCEQ have revealed what appears to be poor flare destruction efficiency by some steam and air assisted flares when operating at turndown condition. Using specialized infrared cameras, unburned hydrocarbons were detected coming from improperly operated flare systems. Despite these findings, the general belief among industry flare experts is that through proper operation of existing flares, the destruction efficiency can be very high even at turndown rates. However, meeting proper destruction efficiency requirements with existing flare systems at low gas flow rates will require more clear and detailed training for operations personnel at these facilities. To help develop this training program, Zeeco performed extensive testing of steam assisted flares at the Zeeco Combustion Research and Test Facility in Broken Arrow, Oklahoma. The testing covered a wide range of operating conditions, specifically focusing on turndown conditions. Testing included emissions plume sampling, as well as real-time data from the same specialized infrared camera currently being used by TCEQ. Computational Fluid Dynamics (CFD) models of the testing were generated and compared to the actual results in order to calibrate the prediction models. The prediction models can then be used to determine if existing flares require any modification to meet current or possible new legislation. The main focus of the testing was aimed at finding and setting the proper operating parameters for steam flares at turndown conditions to allow existing stacks to remain in operation with little or no modification, while still meeting the required destruction and removal efficiencies (DRE). The information obtained from this testing is to be used to develop a training program for flare operators. This paper discusses the approach and results of this testing, as well as recommendations for steam flare operation.