A total of 5,323 appointments from July 1987 to June 1988 was entered into the study. They showed that for the 12 month period the rate of non-attendance was 5%. The non-attender was most likely to have been referred from a GP for a barium meal and be from the age group 16-25. The district the patient came from had little influence. Non-attendance is not affected by the number of appointments given and if compared to the previous year shows a similar monthly pattern. This pattern does not appear to be influenced by the weather but would seem affected to some extent by holiday periods. Reasons for non-attendance were collected on 51 patients. The main reason given was illness. A large group had had no further contact with the referral source. The cost of staffing for non-attendance was considered to be low when compared to the whole budget. The effect of non-attendance on the waiting list is now minimal. Taking this into account the main recommendations are that: a specified senior member of staff should be responsible for the appointment system and inter-departmental liaison; instruction/appointment confirmation sheets should be reviewed; and the counselling service should be improved. Now the population served has been demographically recorded we can look towards providing for its needs.(ABSTRACT TRUNCATED AT 250 WORDS)