Turbulent flame-speed data for premixed flames of methane-air, propane-air, and ethylene-air mixtures stabilized in grid turbulence are reported and discussed. It is shown that turbulence effects on flame speed cannot be correlated fully by the turbulence length scale and rms velocity in the cold flow. Rather there appear to be significant flame-flow turbulence interactions affecting both turbulence level in the reaction zone and measured flame speeds. Results of detailed velocity measurements, including autocorrelations, by laser velocimetry are used to elucidate the nature of these interactions. It is concluded that flame-speed experiments must be designed and conducted to provide sufficient information (e.g., boundary conditions) to allow for reconstruction of the flowfield and these interactions by modelers if the data are to be of value in turbulent combustion model development and evaluation. Nomenclature ? = integral scale of turbulence P = pressure R ( = turbulence Reynolds number, uV/v R x = microscale turbulence Reynolds number, u\/v S L = laminar flame speed S T = turbulent flame speed u = rms velocity, = vw 77 u" = U-U U = instantaneous velocity U = Favre average velocity U = Reynolds average velocity e = Favre average energy dissipation rate X =Tay lor's microscale p = gas density p(r) = autocorrelation coefficient, =