Within the ‘‘impact approximation,’’ calculation of line shapes requires accurate molecular scattering matrices which, in turn, requires solving for molecular collision dynamics on an accurate intermolecular potential energy surface. For some simple, but nontrival molecular systems the interaction potentials are now well determined fromab initioelectronic structure calculations and/or from analysis of spectra of the van der Waals complexes, and it is feasible to solve numerically the coupled channel scattering equations as well. In such cases predicted line widths and shifts are generally in satisfying accord with experiment. Recent results for line widths, line shifts, collisional transfer of intensity in overlapping lines, and Dicke narrowing are presented as examples of current capabilities.