The present research investigates boundary layer heat transfer with very high freestream turbulence, a problem of primary interest in the gas turbine industry where existing boundary layer correlations and codes underpredict local heat transfer rates on first-stage turbine blades and vanes by as much as a factor of three for some engine designs. The problem was studied experimentally by placing a constant-temperature heat transfer surface at various locations in the margin of a turbulent free jet and measuring both the surface heat transfer rate and the turbulence in the free stream. In this experiment, free-stream turbulent fluctuations 20 to 60 percent relative to the mean velocity augment heat transfer 1.8 to 4 times that which would be predicted locally using accepted correlations for turbulent boundary layers at the same Reynolds number.