The pay determining process of CEOs of UK higher education institutions is modelled using three econometric methodologies applied to a large and unique dataset for the academic years 1997/98 through to 2005/06. A gender differential in pay is detected and this differential remains robust across the specifications reported and across higher education sub-sectors. There is evidence that CEOs with industrial work experience and those who have been employed by a higher education body earn more than their counterparts without these attributes. We also find that CEOs are rewarded favourably on the termination of their contracts. There is little evidence that institutional characteristics influence pay after controlling for institution fixed effects. There is only limited evidence that they are rewarded for the 'performance' of the institutions they manage, but are rewarded favourably by increasing the volume of tuition fees. There is some support for 'tournament theory' as an explanation for the determination of CEO pay in this labour market.