In the past 5 years dramatic improvement in the knowledge of the earth's gravity field has been obtained through the use of Geos-3 satellite altimetry data, which has defined the geoid over most of the ocean areas to about 1-m accuracy. In addition, increasing amounts of laser data and improvements in computer programs have permitted the solution for 566 terms in the gravity field. This field was used to compute the orbits of the Geos-3 satellite to an accuracy of 1–2 m in the vertical direction. The combination of satellite tracking data, altimetry data, and gravimetry has yielded 1360 terms in the earth's gravity field. Impressive demonstrations have also been made of the usefulness of satellite to satellite tracking data between high and low satellites for the determination of gravity field. Nevertheless, a number of problems need to be solved in order to provide a gravity field which is sufficiently accurate for the determination and interpretation of oceanographic and geophysical effects and processes. Among these problems are (1) detailed models for effects of atmospheric, solar radiation and earth's albedo on satellite orbits; (2) refinements of effects on station movement, measurement models, and dynamic models owing to crustal motion, polar motion, earth's rotation, solid earth, ocean and atmospheric tides, ionosphere, and troposphere; (3) separation of gravity coefficients where linear combinations of the coefficients are determined by their observed effects on satellite motion; (4) improvements in deterministic and statistical models so that different classes of observational data may be combined without arbitrary weighting of the data; and (5) computational efficiency, since the high cost of the extensive computations often leads to use of inconsistent models for the earth's gravity field for different classes of data. In addition, the usefulness of the altimetry data for gravity field determination is limited by the need to make corrections for topographic effects due to currents, tides, and other oceanographic effects. Surface gravity data are lacking in many areas, and available satellite to satellite tracking data are lacking in both quality and quantity to define medium- and short-wavelength gravity effects to sufficient accuracy.