ESDU Aero W.S.02.03.02 states that at low incidence the lift-dependent drag coefficient of a wing in supersonic flow can be written as the product of the incidence and the lift coefficient minus the leading-edge suction force coefficient. For sharp supersonic leading edges the suction force is zero but for wings on which the leading edge is wholly or partly subsonic the suction force must be taken into account. Using linearised theory a graph presents a factor as a function of Mach number and leading-edge sweep; the suction force coefficient is then the product of that factor, the aspect ratio and the incidence squared. However, it is noted that there is a maximum suction corresponding to a vacuum over the complete front part of the wing. In practice only a small fraction of that maximum occurs. To assess that fraction, the local section normal to the leading edge is considered as the forward part of an elliptic cylinder; for that section the theoretically-derived ratio of the average local suction per unit thickness to the maximum local suction is plotted against thickness/chord ratio. Finally a graph plots as a function of Mach number an empirical factor that relates the local maximum suction to the theoretical maximum due to vacuum. It is noted that the data represent the trends in lift dependent drag well but that the value of the suction force coefficient is over-predicted. A worked example illustrates the use of the curves.