Radar wind profilers detect scatter from clear-air refractive-index irregularities. The Doppler shift calculated from the time series of the backscattered signal provides an estimate of the radial velocity of the air within the radar's resolution volume. It is known that there are quite a number of effects that can give rise to intrinsic biases in these radial-velocity estimates. It is extremely difficult to unambiguously identify these biases on the basis of observations since it is impossible to independently probe the three-dimensional, time-dependent fine structure of the relevant atmospheric variables (wind and refractive index) within the radar's resolution volume down to the relevant spatial and temporal scales associated with the scattering process itself. In this paper the large-eddy simulation (LES) technique is used to generate, with high spatial and temporal resolution (vertical resolution, 8 m; horizontal resolution, 16 m; time step, 0.8 s), data of atmospheric variables within a convective boundary layer. The LES data that are used in this study to not have the spatiotemporal resolution that would be needed for a deterministic, full solution of the scattering integral. However, a parameterized model of the scattering integral can be constructed on the basis of Kolmogorov-scaling arguments and a random-phase model. This model can be numerically solved on the basis of LES data. For the first time, LES-generated fields of virtual temperature, humidity, and three-dimensional wind are used to construct time series of the real and imaginary parts of wind-profiler signals. The simulated wind-profiler signals are processed to derive Doppler velocities, and the simulated Doppler velocities are compared with the (simulated) meteorological in situ truth provided directly by the LES wind-velocity data set. Examples of simulation results and the potential of the LES-based, wind-profiler simulation technique as well as some limitations are discussed. ¿ 1999 American Geophysical Union