The funding for research activities in English universities from 1993/94 on is based on a formula allocation which has many apparent anomalies: variations in the standardised sums available per discipline bear little relationship to interdisciplinary differences in research costs. Using data released by the Higher Education Funding Council for England, I seek to understand how those differences have come about. Only 43% of the variation could be accounted for statistically by a simple classification of disciplines into clinical, science, and social science/humanities. A further 17% was associated with four other factors reflecting institutional differences in aspects of the disciplines which could have been taken into account when the allocation formula was determined-but which were not. The conclusion is that a much fairer system can and must replace that currently operated.