Evidence of variability in degree results in UK universities from 1973 to1993 is presented for eight subjects. This variability is not random. Thereis a clear tendency in all subjects for certain universities to awardpercentages of Good degrees (i.e. Firsts and Upper Seconds) consistentlyabove or below the corresponding national average. These differences maypartly reflect entry qualifications with universities acceptingbetter-qualified students awarding more Good degrees. Other plausibleexplanations may, however, be offered including institutional differences inteaching and assessment practices and in the definition of class boundaries.It is important, in the context of the current debate about 'standards', totest these explanations by conducting qualitative research within thedepartments responsible for the delivery of teaching and for the assessmentof students.