Accessory minerals play a crucial role in the petrogenesis of granitic rocks. They control the response of isotopic systems (U–Pb, Lu–Hf, Sm–Nd) and rule the geochemical variation of many important trace and minor elements during partial melting and fractional crystallization. The experimental effort in the past thirty years has resulted in the formulation of saturation models with manifold applications in igneous geochemistry. Numerical simulations show that the zircon saturation thermometer is rather insensitive to analytical errors, the presence of a minor inherited component, as well as moderate accumulation of feldspars and some ferromagnesian minerals (Ca-poor pyroxenes, less so olivine). However addition of even minor cumulus Ca-pyroxene or Ca-amphibole would quickly render the calculated temperatures too low. Apatite saturation thermometry is a poor tool for felsic metaluminous rocks, being oversensitive to errors in the phosphorus determination as well as the presence of extraneous apatite. Even stronger element of uncertainty is added for peraluminous lithologies, when the increased apatite solubility is inadequately accounted for by the current models and where other phosphorus-bearing minerals, most importantly feldspars and monazite, come into play. The newly developed software Saturnin (http://www.gla.ac.uk/gcdkit/saturnin) written in the freeware R language performs otherwise tedious calculations of zircon, monazite and apatite saturation in igneous rocks. While the Windows users would probably find it easier to access the programme as a part of the larger package GCDkit (http://www.gla.ac.uk/gcdkit), on Macintosh and Linux it can be used as a stand-alone application.