Heat flux measurement in a hypersonic wind tunnel is one of indispensable requirements for designing the next generation re-entry vehicle. Heat flux is usually calculated from a temporal change in temperature on the model surface. Temperature-Sensitive Paint (TSP) is a global temperature measurement technique based on photochemical reaction. This technique was applied to a hypersonic shock tunnel, but there existed uncertainties caused by TSP layer thickness. In this study, the relationship between TSP layer thickness and measurement accuracy in heat flux was studied by experiment. TSP composed of Ru(phen) 32+ as luminophore and polyacrylic acid as binder was applied on a cylinder model made of glass ceramics. The thickness of TSP layer was varied from 0.2 to 3 μm. The models were tested at Mach 10 in the JAXA 0.44-m Hypersonic Shock Tunnel and heat flux caused by aerodynamic heating on the model surface was measured. A comparison between TSP data and conventional thermocouple data showed that the measurement error changed with the layer thickness In nonlinear fashion. In addition, the difference between TSP and thermocouple measurement was larger than theoretical prediction. For TSP layer thickness less than 0.5μm, the accuracy in heat flux measurement by TSP was within the uncertainty of conventional thin thermocouple measurement. To demonstrate the capability of TSP with the optimized thickness, tests using a 3-dimensional wing-body model have been conducted and the complicated heat flux pattern caused by shock-wave/shock-wave interaction has been observed.