The first occurrence of immiscibility in magmas appears to be most important in the magmatic-hydrothermal transition, and thus studies of magmatic immiscibility should be primarily directed towards recognition of coexisting silicate melt and essentially non-silicate liquids and fluids (aqueous, carbonic and sulphide). However, immiscible phase separation during decompression, cooling and crystallization of magmas is an inherently fugitive phenomenon. The only remaining evidence of this process and the closest approximation of natural immiscible magmatic liquids and vapours can be provided by melt and fluid inclusions trapped in silicate glasses and magmatic phenocrysts. Such inclusions are often used as a natural experimental laboratory to model the process of exsolution and the compositions of volatile-rich phases from a wide range of terrestrial magmas. In this paper several examples from recent research on melt and fluid inclusions are used to demonstrate the significance of naturally occurring immiscibility in understanding some large-scale magma chamber processes, such as degassing and partitioning of metals. 2011 Blackwell Publishing Ltd.