Non-attenders for five years or longer (77 men and 27 women; i.e. 1.9 per cent of the practice) in a general practice of 5,750 patients were identified and paired with controls in the same practice. A survey by questionnaire was carried out.The conclusions were:(1) A higher proportion of non-attenders are self-employed,(2) A lower proportion of them had retired than their attending counterparts,(3) They are spread over the whole age-range, but with a higher concentration in the ages 35 and over,(4) They are evenly spread over the social classes,(5) They seldom seek other sources of medical advice or treatment,(6) They are significantly slimmer than their fellows,(7) Their non-attendance does not, on the whole, cloak serious but remediable illness.