Comets are the only large solar system bodies where nongravitational forces directly affect their dynamic motions. Their approach to within a few AU of the Sun initiates the vaporization of nucleus ices, and the resulting rocket-like effects either add to or subtract from the comet’s orbital energy; the sign of the energy change depends upon the comet’s rotation direction and its spin pole orientation. The cometary outgassing phenomena have generally been modeled by assuming a rapidly rotating nucleus of water ice that outgasses symmetrically with respect to perihelion. Although this nongravitational acceleration model has been quite successful in providing accurate orbits and ephemerides, several comets exhibit water production rates and visual light curves that are noticeably asymmetric with respect to perihelion. New asymmetric models are being developed that attempt to represent more closely the cometary outgassing phenomena. For the same comet, derived nongravitational parameters can differ widely, depending upon which model is used to fit the astrometric data. The uncertainties in the data and in the nongravitational acceleration model prevent realistic extrapolations of these objects’ motion beyond a few hundred years, particularly if close planetary encounters are involved. Accurate orbits, ephemerides and efforts to model the nongravitational effects ultimately depend upon the quality of the astrometric data. Using a combination of long-focus telescopes, charge coupled device (CCD) detectors, microdensitometer reductions and modern star catalogs, cometary astrometric data can be generated that are accurate to the sub arcsecond level. While occultation, spacecraft, and radar observations can provide powerful astrometric data when available, it is still the ground-based optical observations that must provide the vast majority of data for cometary astrometry in the foreseeable future.