We investigate the geoeffectiveness of the 3 WIND magnetic clouds of October 1995 (OCT95), May 1996 (MAY96) and January 1997 (JAN97). We use "Geoeffectiveness" in its traditional sense as a collective term for magnetospheric and ionospheric disturbances including storm and substorm activity and other large-scale, global effects such as those due to untypically large solar wind dynamic pressures and variations thereof which are rapid and of large amplitude. The interplanetary signatures of all three clouds bear a strong resemblance to one another [in temporal profile but not in amplitude] and all three formed part of compound streams, with a faster stream overtaking each. The comparison envisaged is thus meaningful. Using a many-station determination of the Dst storm index, corrected for the contribution of magnetopause currents, and using the threshold Dst < -100 nT for major storms, we find that OCT95 caused a major storm, while JAN97 and MAY96 caused only moderate storms. Ring current activity is substantially enhanced within minutes of cloud onset in all 3 cases, and subsides at the end of cloud passage, only to resume at a diminshed level due to fluctuating Bz < 0 fields and large dynamic pressure changes in the following fast solar wind stream. We monitor substorm activity by a multi-station determination of the AE and AL indices. In addition, we compare the geomagnetic effects of the clouds around a latitudinal circle which would usually be classed as sub-auroral, particularly the effect in the local midnight sector. We find that substorm activity and especially its sub-auroral effect is comparable on OCT95 and JAN97, but is much reduced on MAY96. The Poynting flux into the magnetosphere, a good measure of geoeffectivenss, is as follows: OCT95:JAN97:MAY96=8:4:1. Other large-scale effects which occur during cloud passage will be compared and discussed in detail. In summary, we find that in most aspects the OCT95 is the most geoeffective of the three clouds, and MAY96 the weakest. We elaborate on the causes of this in the presentation.