An exploratory study of the laminar, transitional and turbulent boundary layer on a 5 deg. half-angle cone in hypervelocity flow was conducted in the high-enthalpy shock tunnel T5 by measurement of the heat flux distribution and by qualitative flow visualization. A novel flow visualization technique using sodium seeding to increase the sensitivity of conventional interferometric techniques by resonant enhancement of the refractivity of the medium was developed to study the boundary layer structure. The experiments were designed to cover a large range of specific reservoir enthalpy, ranging from the perfect-gas regime to the range where significant oxygen and some nitrogen dissociation and recombination effects may be expected in the boundary layer. The presence of atomic species is due to the combined effect of nozzle freezing and frictional heating in the boundary layer. In the laminar regime and in the latter range, the following effects were found to be present:​ At the same nominal conditions, heat flux levels are higher in air than in nitrogen because of a larger heat release from oxygen recombination at the wall. By varying the reservoir specific enthalpy in air and nitrogen, and from measurements in carbon dioxide, it was found that real-gas effects stabilize the boundary layer. If the transition Reynolds number is renormalized by evaluating it at the reference temperature, the data for a given gas becomes correlated in a plot against reservoir enthalpy. Increasing enthalpy stabilizes the flow. The stabilizing effect is stronger with gases whose lowest activation energy is low. This behavior is opposite to the prediction made by the linear stability theory regarding the second linear mode of instability. The linear stability theory predicts, however, that real-gas effects stabilize the Tollmien-Schlichting mode. Flow visualization results suggest that the dominant instability mode in the present experiments was the Tollmien-Schlichting mode. Finally, the flow visualization pictures show structures that are not qualitatively different from those of an incompressible turbulent boundary layer, but they do not indicate if real-gas effects change significantly the structure of the turbulent boundary layer. The heat transfer measurements compare well with semi-empirical predictions.