Five hundred and twenty-nine patients attended five different clinics supervised in three different ways in two adjacent, but socially different, health districts. Performance was apparently better in the clinics supervised by the consultant personally or by specialised junior staff. Non-atopic asthma appeared to be more frequent in clinics attended by patients with a social advantage, where it also appeared to be more easily controlled. This difference in control was less apparent in atopic asthma. These differences and deliberate selection probably account for the differences in performance between the clinics. Prescription policy appeared to be uniform throughout the clinics, though there was a tendency for drugs to be given at a lower level of pulmonary function in those clinics where performance was generally worse. Five per cent of all patients failed to achieve satisfactory treatment after a whole year's observation. This was not influenced by supervision or social group. Analysis of the pulmonary function of patients for whom drugs had been prescribed suggested that there was possible under-use of disodium cromoglycate as opposed to inhaled corticosteroids. The exercise proved a rapid and useful assessment of the performance of the clinics supervising asthma.