The validation and specification of the capacity of spaceborne radar altimeters to estimate marine wind speed and ocean significant wave height are dependent upon comparisons of these quantities as estimated by radar altimeters and conventional in situ sensors mounted aboard surface buoys. Two important questions are associated with these comparisons. First, what are the expected differences between buoy and altimeter estimates of wind speed and significant wave height? Second, given a knowledge of these expected differences and a finite number of buoy--altimeter comparisons, what conclusions can be reasonably drawn about the capacity of an altimeter to estimate wind speed and significant wave height? In this paper we outline and quantify the expected differences between buoy and altimeter estimates of wind speed and significant wave height. These differences are categorized as those associated purely with the buoy, purely with the altimeter, or the disparate manner in which buoys and altimeters sample the spatially and temporally varying wind and wave field. Based on these expected differences, statistical tests are given to validate and specify altimeter performance. In addition, statistical approaches to discriminating between candidate algorithms for converting the return pulse characteristics of a radar altimeter into wind speed estimates are discussed. ¿ American Geophysical Union 1988