Eighty five general practitioners in the Lothian region recorded information on all surgery consultations on one day in every 15 for a year. On the basis of their mean consultation times with patients the working styles of the general practitioners were described as 'faster' (n = 24), 'intermediate' (n = 40) or 'slower' (n = 21). The 21,707 consultations which they carried out over this period were defined as 'short' (five minutes or less), 'medium' (six to nine minutes) or 'long' (10 minutes or more). Independent of doctor style, 'long' consultations as against 'short' consultations were associated with the doctor: (1) dealing with more of the psychosocial problems which had been recognized and were relevant to the patient's care; (2) dealing with more of the long term health problems which had been recognized as relevant; and (3) carrying out more health promotion in the consultation. Patients also reported greater satisfaction with longer consultations. The ratio of long:short consultations was found to be 0.28:1 for 'faster' doctors as against 2.3:1 for 'slower' doctors. When doctors in either group had more heavily booked surgeries or were running late, the long:short consultation ratio fell, in some cases by over 50%. This paper suggests that the ratio of long to short consultation length for a general practitioner might become the basis of a simply proxy measure of quality of care; and that its use might help monitor the effect of recent and proposed changes in the way in which general practice care is delivered.