National allocation of resources to regional health authorities and by them to districts is now determined by a weighted capitation formula. The national formula was derived from regression analysis, with hospital utilisation as an index of need for health care--a method which has fundamental limitations. This paper argues that the search for an empirically based resource allocation formula of high precision in the name of promotion of equity is largely fruitless given the impossibility of measuring the true need for, and costs of, providing health care, especially with the limited data available. The inclusion of measures of social deprivation is also poorly thought out. The availability of data from the 1991 census, which included a question regarding long-standing illness, together with the intention of the Department of Health to review the weighted capitation formula using this information may stimulate much work but little light. It is essential that the impact of resource allocation formulas is justifiable on grounds other than the composition of any particular formula.