The biological and technical hurdles confronting the development of new diagnostic tools are manifestly great in leprosy. Leprosy diagnosis in the field, for control as well as for research purposes, will have to remain, for the time being, predominantly clinical. It is important that the significance and relevance of diagnosing so-called 'early' lesions must be viewed in the context of the objectives of leprosy control. All evidence suggests that the majority of these 'cases' are not likely to progress in the individual, nor is there any evidence that these 'cases' are of any importance for the transmission of the disease. At its best, such 'cases' may be a sign of temporary infection which will disappear spontaneously within a few months without leaving any residual signs. It is more likely that the majority of these are not cases of leprosy. In making such a diagnosis its ethical implications on the individual, his family and the local society must be considered, given the intense social stigma this diagnosis will generate; besides the cost of treatment and the potential risk of serious side-effects due to treatment. It may be justifiable to have a broad definition if the incidence of leprosy is still high and all reported cases are to be accepted as valid. But once the backlog of cases has been detected and, as a result of control efforts, the number of new cases begins to decrease, there is a need to narrow the case definition for national statistics.(ABSTRACT TRUNCATED AT 250 WORDS)