Space-based instrumentation is often capable of generating more data than the spacecraft telecommunications downlink can handle. This means that potentially valuable data has to be discarded before it can be looked at, or that expensive instrumentation is not working to full capacity. Data compression can help to increase the effective productivity of space-based instrumentation. However data compression is not free: it requires additional on-board processing to perform the compression and may result in some degradation of image quality. Synthetic Aperture Radar (SAR) instruments with high resolution and all weather, day and night capability produce very large volumes of data on-board a satellite. This paper looks at several different methods of raw SAR data compression. These methods were implemented and tested on a set of raw SAR data. The effects of the different forms of data compression on final SAR image quality were assessed. The results of this evaluation are presented here together with comparable results from other groups investigating raw SAR data compression. Data compression techniques which operate directly on the raw SAR data, on partially processed data or on image data are considered.