In-cloud motion vector retrieval is of great interest in several atmospheric science research fields. Short time extrapolation of radar data (precipitation nowcasting), assimilation into numerical weather prediction models, study of atmospheric circulation, as well as reference scenarios for future satellite missions, are glaring example where the knowledge of in-cloud motion vectors can play a relevant role. In this work, a dataset of nearly one-year and half of measurements collected by ground-based weather radars over the Italian peninsula, is used to perform the reconstruction of horizontal in-cloud rain motion vectors (RMVs) using both optical flow-based solutions from literature and an innovative extension of the multiple Doppler solution that make use of mosaicked Doppler radar data. The outcomes of the techniques imple-mented are analyzed in terms of reference Doppler measurements, reanalysis wind fields from ERA5 and evaluating the impact of the RMVs in a semilagrangian precipitation nowcasting framework. To the author knowledge, this is the first attempt in quantitatively evaluating RMVs. Results show that the use of Doppler information enhances the dynamic range of the retrieved RMV intensity with respect to optical flow-based solutions giving a better agreement with the ERA5 too. In terms of precipitation nowcasting, the use of Doppler-driven RMV does not give significant improvements due to gradients shown by RMV intensity when constrained with the measured Doppler.