We revisit the problem of computing the backscattering coefficient based on the measurement of scattering at one angle in the back direction. Our approach uses theory and new observations of the volume scattering function (VSF) to evaluate the choice of angle used to estimate bb. We add to previous studies by explicitly treating the molecular backscattering of water (bbw) and its contribution to the VSF shape and to bb. We find that there are two reasons for the tight correlation between observed scattering near 120$^\circ$ and the backscattering coefficient reported by Oishi (Appl. Opt. 29, 4658, 1990), namely, that 1. the shape of the VSF of particles normalized to the backscattering does not vary much near that angle for particle assemblages of differing optical properties and size, and 2. the ratio of the VSF to the backscattering is not sensitive to the contribution by water near this angle. We provide a method to correct for the water contribution to backscattering when single-angle measurements are used in the back direction, for angles spanning from near 90$^\circ$ to 160$^\circ$, that should provide improved estimates of the backscattering coefficient.