Measurements and predictions of the X-33 turbulent aeroheating environment have been performed at Mach 6, perfect-gas air conditions. The purpose of this investigation was to compare measured turbulent aeroheating levels on smooth models, models with discrete trips, and models with arrays of bowed panels (which simulate bowed thermal protections system tiles) with each other and with predictions from two Navier-Stokes codes, LAURA and GASP. The wind tunnel testing was conducted at free stream Reynolds numbers based on length of 1.8 x 10(exp 6) to 6.1 x 10(exp 6) on 0.0132 scale X-33 models at a = 40-deg. Turbulent flow was produced by the discrete trips and by the bowed panels at ill but the lowest Reynolds number, but turbulent flow on the smooth model was produced only at the highest Reynolds number. Turbulent aeroheating levels on each of the three model types were measured using global phosphor thermography and were found to agree to within .he estimated uncertainty (plus or minus 15%) of the experiment. Computations were performed at the wind tunnel free stream conditions using both codes. Turbulent aeroheating levels predicted using the LAURA code were generally 5%-10% lower than those from GASP, although both sets of predictions fell within the experimental accuracy of the wind tunnel data.Hollis, Brian R. and Horvath, Thomas J. and Berry, Scott A.Langley Research CenterX-33 REUSABLE LAUNCH VEHICLE; AERODYNAMIC HEATING; HYPERSONIC SPEED; TURBULENT FLOW; SCALE MODELS; WIND TUNNEL TESTS; COMPUTATIONAL FLUID DYNAMICS; NAVIER-STOKES EQUATION; ALGORITHMS; BOUNDARY LAYER TRANSITION; REYNOLDS NUMBER; PANELS; THERMOGRAPHY; GLOBAL AIR SAMPLING PROGRAM; FREE FLOW...