Over the past two decades epidemiologists interested in chronic illness have tended to concen trate on conditions which are lethal, and they have generally directed more attention to the elucidation of the causes of such diseases than to their social consequences. Although this work has been of great importance, one result is that we know less than we should about the burden on society of many diseases, especially those with a low case fatality rate such as the arthritides, varicosities, and even diabetes. To date one of the few large-scale sources of information of this kind is in the United States where the National Health Examination Survey (1966a and b, 1968) has shown that the prevalence of arthritic conditions is extremely high, although it has not included data on impairment and disability in objectively ascertained cases. In contrast, the US National Health Interview Survey (1960, 1964, 1969, 1971), which has had to depend on the respon dent's word for ascertainment of diagnosis, has collected data on the extent of limitations of activity. Throughout that country arthritis and rheumatism, which are treated in the National Health Survey as a single category, were by far the most serious cause of limited mobility, and in many areas they are also the most important chronic cause of limited activity (US National Health Survey, 1971). If arthritis and rheumatism are taken in combination with impair ments of the spine, hips, and legs, they have con sistently, between 1961 and 1967, accounted for 29% of all cases of limited activity and in this respect far outstrip in importance any other group of chronic diseases. Over the same seven-year period the next most important conditions have been those of the cardiovascular system, yet 'heart conditions' and 'hypertension without heart involve ment' taken together have accounted for only 22% to 23 % of the cases of chronically limited activity (US National Health Survey, 1968). In 1967 some 3,250,000 civilians who were not in hospitals or other institutions suffered limitation of activity because of these two conditions and in well over half of these there was serious loss of time at work (US National Health Survey, 1971). The communal problem of arthritis and rheumatism is accentuated by the fact that the burden imposed by them is heaviest among the poorest members of the com munity. In Britain the General Household Survey is collecting data which are similar to those from the US Interview Survey (1964, 1969, 1971). This paper presents some information about the interrelationship between impairment and disability and arthritic conditions in a group of adults from the general population of New Haven, Connecticut, and combines objective with interview data.