Introduction In countries where mortality statistics are reason ably complete but where morbidity statistics are not, there is a danger that variations in death rates will be accepted too readily as indicative similar varia tions in sickness rates. The errors such acceptance are obviously small for which are usually fatal, and the greatest difficulties arise in deducing secular trends morbidity from mortality data with decreasing fatality rates. Northern Ireland is such a country and respiratory tubercu losis is such a disease. This particular combination and country is probably studied more easily than others because in Northern Ireland all the tuberculosis services, for the whole population about 1 4 million, are controlled by a single authority, the Northern Ireland Tuberculosis Authority. The purpose this paper is to examine the existing morbidity and mortality data in an attempt to assess their value as measures the success the fight against respiratory tuberculosis. The crude death rate attributed to respiratory tuberculosis in Northern Ireland fell from 128 to 14 deaths per 100,000 the population between 1922 and 1954. During the first 24 years the rate was reduced by 50 per cent., and during the last 8 by more than 75 per cent. A full appreciation this trend cannot be made without reference to the age and sex distribution the mortality, because the changes have not been uniform in every age and sex group; indeed in some, notably men aged 55 years and over, the rates have increased (Cheeseman, 1952). It is more difficult to obtain comparable morbidity rates. Mortality rates are little value as an in direct measure the trend morbidity because they are influenced not only by the volume but also by increasing success in the treatment infected persons. For the community as a whole, few morbidity data are available prior to the estab lishment the Northern Ireland Tuberculosis Authority in 1946. To put their notification records into proper perspective, the impact the Authority's work on the tuberculosis problem must be appre ciated. The Authority was instituted with a unique responsibility, for the whole country, of securing, in co-operation with sanitary and other local authori ties, the prevention and more effective treatment tuberculosis and kindred diseases (Northern Ire land Tuberculosis Authority, 1948). The resources inherited in 1946 were meagre; they consisted some 500 beds in five hospitals, three chest clinics with x-ray facilities, a static mass miniature radiography unit, and some forty poorly equipped dispensary premises. The development these resources and the new facilities provided since 1947 have been described elsewhere (Brit. med. J., 1954; Elder, 1953), and the present situation may here be summarized briefly. About 970 beds are now available in the Authority's hospitals and 700 in other hospitals. There are 28 chest clinics, all but six being attached to hospitals with full diag nostic facilities. Two mobile mass radiography units have supplemented the existing static unit and a B.C.G. vaccination programme is in operation. The service is medically staffed on an area basis with a consultant and two or more chest physicians, or assistant chest physicians, in each area. Notifications respiratory tuberculosis under the Authority's scheme are available by age, sex, and state disease classification for each year since 1948. In 1952, it was decided to exclude patients with active primary tuberculosis from the published notifications (Kelly, 1955), but the relevant details such patients from 1952 onwards have been made available to me (Harvey, 1955), and in this report such patients are included in all references to notifi cations. The Authority uses the classification 115