The calibrations of four airspeed systems installed in a North American F-86A airplane have been determined in flight at Mach numbers up to 1.04 by the NACA radar-phototheodolite method. The variation of the static-pressure error per unit indicated impact pressure is presented for three systems typical of those currently in use in flight research, a nose boom and two different wing-tip booms, and for the standard service system installed in the airplane. A limited amount of information on the effect of airplane normal-force coefficient on the static-pressure error is included. The results are compared with available theory and with results from wind-tunnel tests of the airspeed heads alone. Of the systems investigated, a nose-boom installation was found to be most suitable for research use at transonic and low supersonic speeds because it provided the greatest sensitivity of the indicated Mach number to a unit change in true Mach number at very high subsonic speeds, and because it was least sensitive to changes in airplane normal-force coefficient. The static-pressure error of the nose-boom system was small and constant above a Mach number of 1.03 after passage of the fuselage bow shock wave over the airspeed head.