The increasing use of CCD’s and IRCCD’s to detect stellar companions, either by direct imaging or speckle techniques, raises the issue of the fraction of companions whose instantaneous, projected separations lie in the range corresponding to the geometric limits of the detector. The issue is of especial importance to recent surveys employing IRCCD detectors of relatively very small area (projected onto the sky). The distribution of observed separations reflects not only that of orbital semi-major axes, but also distributions of distances to binary systems, orbital eccentricities, tilt of orbital planes with respect to the plane of the sky, and of arguments of periastron. Proper inference of, e.g., binary multiplicity requires estimates of completeness that should incorporate contributions from all these factors.