Whether the phenomenon known as a flux transfer event (FTE) is caused by a special process acting as the magnetopause such as transient reconnection or whether it is simply a double crossing of the magnetopause caused by a pressure pulse has become controversial. Theoretical investigations have shown that there are some similarities in the flow and field perturbation between a magnetopause surface wave and a convecting cylindrical structure. This result by itself does not distinguish the two possibilities. We investigate the observed differences between FTEs and surface waves. In general, a realistic magnetopause surface wave has a much longer wavelength than the displacement, which is different from some models of surface waves. The scales of an FTE along the convection direction and normal to the magnetopause are of the same order. Spectral analysis of the solar wind and upstream waves shows no evidence of pressure variations of the duration and intensity to generate the magnetopause surface wave of the required wavelength of few Earth radii and a displacement of an Earth radius. The difference between a surface wave and a series of FTEs can be easily recognized in the profiles of the magnetic field and plasma parameters. The FTEs look like a series of pulses, whereas multiple crossings of the magnetopause occur with a changing ratio of the duration of magnetosheath and magnetospheric entry. The timing difference between two spacecraft during magnetopause crossings is significantly greater than the timing difference of FTEs. For typical magnetopause crossings the perturbations in the normal component and magnitude of the magnetic field is not more than background magnetosheath fluctuations; but those for FTEs are often much greater. The reinterpretation of FTEs as surface waves arose without a quantitative assessment of the observed perturbations of the magnetopause and quantitative effects of proposed sources. When such a quantitative assessment is performed, it is clear that FTEs are not simply magnetopause crossings. ¿ American Geophysical Union 1994