Multi-channel SAR systems allow high-resolution Imaging of a wide swath and represent an attractive solution for present and next-generation SAR mission concepts. However, their operation requires the downlink of a huge amount of data: in addition to the intrinsic data equirement related to resolution and swath, in fact, a pulse repetition frequency (PRF) higher than the processed Doppler bandwidth is usually selected. A convenient data reduction strategy is therefore proposed, which exploits the correlation between samples by performing variable-bit quantization of the data after a block transformation of the multichannel SAR data, e.g., an FFT. Simulations show that this allows reducing the data volume while keeping the same performance at the cost of a modest onboard processing effort.