The Mesozoic, perhaps representing the longest period of warmth during Phanerozoic Earth history, contains in general sparse and frequently equivocal evidence for polar ice. Although this time is undoubtedly punctuated by oscillations in climate, whether sufficient to lead to polar cooling and the formation of polar ice (an inviting mechanism to account for faunal and floral distribution patterns and large scale sea level change), has been widely debated. Mesozoic evidence for glacial conditions includes abraded rock surfaces, generally unsorted stone-rich beds and the presence of dropstones rafted by ice within a finer-grained host sediment. Faunal and floral evidence has also been utilised to determine the presence or absence of cold or sub-freezing polar conditions, together with more indirect evidence for glacial conditions derived from General Circulation Model (GCM) simulations of climate, the analysis of clay mineral distributions, glendonite abundance, and palaeontological and sedimentological evidence for globally synchronous sea level change. The extent of possible glacial environments during the Mesozoic has been established by plotting published reports of glacial sediments on palaeogeographic reconstructions. The general clustering of evidence at high palaeolatitudes suggests that the extent of polar ice during the Mesozoic is likely to have been approximately one third the size of the present day. Based on such evidence a number of episodes of cold or sub-freezing polar climates during the Bajocian-Bathonian, Tithonian/Volgian, Valanginian and Aptian are recognised. Evidence exists possibly also for a cold episode during the early Jurassic (?Pliensbachian) although poorly constrained owing to limited biostratigraphical control. The longevity of these events may be represented by a `cold snap' within a stage, although undoubtedly the possible `smearing' of ages may have had the effect of lengthening and hence promoting the importance of these proposed events. The climate regime of the Earth during these times may be hypothesised to be characterised by a relatively steep pole-to-equator temperature gradient where low-latitude regions are as warm or warmer than today. The evidence to support such reasoning includes sharp increases in the areal distribution of both glendonites and deposits with affinities to glacial tillites and dropstones for these times. Coincidental falls of sea level, arid events and an increased bipolarity of faunas at these times imparts further confidence that these events were of a magnitude to effect the Earth as a whole.