Three dimensional velocities are required to constrain halo shape total mass, but proper motions (tangential velocities; hereafter PMs) are notoriously difficult to measure. I will present results from our efforts in precision astrometric techniques. First, HST programs aimed at measuring the PMs of the LMC & SMC have now achieved sufficient precision to go beyond the absolute motions. By mapping variations in PM across the face of the LMC, it has been possible to measure its PM rotation field & rotation curve, the first time this has been possible for any galaxy. Second, I will present efforts to measure the PMs of tidal streams in the halo using both a space-based & a ground-based approach. The wide-field, ground-based imagers typically cannot give as good per-star accuracy as HST, but have the advantage of statistics & efficiency: SDSS already exists as a first epoch, & the wide field of view (FOV) plus the use of a big telescope gives good performance in a “stars per hour exposure time” metric. However, for confident separation of tidal debris from Milky Way foreground/background, additional information such as a matched filter must typically be used. HST gives very high accuracies per star, & therefore tidal debris can typically be separated from Milky Way stars by means of PMs alone. However, this is only possible for small samples of stars because of the small FOV’s of these instruments. Thus, a combined approach is warranted.