We examine the statistical estimation error for the dual coplanar beam momentum flux estimator in the presence of geophysical noise. Included in our derivation are the effects of measurement noise and finite spatial correlation of the wind fluctuations. We show that there is an optimal beam separation that minimizes the statistical estimation error at a value nearly equal to the irreducible error observed by an ideal anemometer measuring u and w at a point. For typical mesopause parameters this beam separation is approximately ¿13¿. The optimal angle is dependent in part on the ratio of vertical to horizontal wind variances and is therefore variable. For optimal momentum flux measurements a system, radar or lidar, with steerable beams becomes necessary. ¿ 2000 American Geophysical Union