We audited the case records of 493 new patients referred to a chest clinic to determine for eight consultants and five middle-grade staff the average number of investigations performed, the follow-up rate and prolixity (the length of the letter written to the general practitioner). The outcome of the consultation was assessed by a questionnaire administered to the patients' general practitioners. Practitioner mean investigation rates varied from 0.1 to 8.7 investigations per patient, follow-up rate from 45 to 100% and average prolixity from 53 to 200mm. The prolixity of the eight consultants was significantly related to their investigation and follow-up rates. Consultation outcome was at least satisfactory for 97.4% of patients. We conclude that considerable potential exists for savings of laboratory, clinic and secretarial time and costs by reducing investigation and follow-up rates and writing shorter letters to general practitioners.