It is shown that the assumption that sodium can be used as a passive tracer for studying atmospheric waves in the region above 90 km is not always valid. An analysis of wave structures observed in the sodium layer by lidar shows that the associated oscillations in the mixing ratio of sodium are frequently too large to be produced by vertical displacements in the presence of mixing ratio gradients. On some occasions these large amplitude oscillations appear to be associated with the formation of sporadic sodium layers, and it is suggested that both phenomena are most probably related to the windshear concentration of sodium ions. ¿ 1998 American Geophysical Union