This paper develops the feasibility of extracting useful dynamical information from chemiluminescence polarization measurements carried out under beam-gas conditions rather than crossed beam conditions in order to achieve higher signal levels. It extends the previous work of Prisant et al., 1981, J. chem. Phys., 55, 2222, to include supersonic beam conditions. Classical analysis gives a simple relationship between the polarization index, R, and the product rotational alignment, (P~(J' . ~)). The cylindrical symmetry of the beam-gas configuration allows factorization of the alignment into dynamical and kinematic factors. The kinematic factor takes into account the fluorescence depolarization which arises from the spread of relative velocity vectors about the beam axis. Its dependence on the mass and velocity of the collision partners is derived as a function of two dimensionless parameters, p and ~, which characterize all beam-gas configurations. The results show that for much of the parameter space, the extent of depolarization is less than I0 per cent.The spread in relative velocity vectors also leads to a degradation of the collision energy resolution. A simple analytical expression is obtained for the mean collision energy and the energy resolution is also calculated as a function of the beam-gas parameters, p and ~. The two sets of results provide a useful guide to experimentalists choosing suitable systems for study under beam-gas conditions.Measurements of the chemiluminescence polarization as a function of collision energy, obtained using a rotor accelerated beam technique, for the kinematically constrained reaction :Xe*(~P~,0) + HCI -+XeCI* + H -~Xe + C1 + H + h v are found to be in good accord with theory.