The influence of long ocean waves on the determination of surface winds from the scatterometer of the European remote sensing satellites (ERS-1/2) is investigated theoretically. The scatterometer operates at 5.3 GHz with vertical polarization (VV). Presently operational algorithms for wind retrieval are based on empirical scatterometer models, which describe the normalized radar cross-section sigma0 in terms of wind speed and direction. These models account for both resonant backscattering from small ripples and their modulation by longer surface waves (wind sea). By making use of the composite surface scattering theory, the portion of sigma0 due to long-wave modulation is quantified. The decomposition of sigma0 allows the investigation of two questions. (1) Are sigma0-models derived for the scatterometer applicable for the synthetic aperture radar (SAR)? On ERS-1/2 both measurements are performed by the same instrument. However, the illuminated areas differ considerably. (2) May the presence of swell distort the retrieval of wind direction? Empirical sigma0-models account for the wind sea but not for swell, which is an occasional phenomenon with a direction independent of the local wind. Under certain conditions, significant influences have been found: (a) underestimation of wind speed with SAR data of up to 20%, and (b) errors in wind direction due to the presence of some 40°.