This study was conducted to assess the usefulness of IgM anti-PGLI antibody test, for the diagnosis of leprosy and of the subclinical infection among contact population. Even if the specificity (97.7%), sensitivity (98% for the multibacillary, but 36% for the paucibacillary) and efficiency (97.7%) of the test for the diagnosis of patients are good, its positive predictive value which is the proportion of true patients among the seropositive subjects detected in the population was very low (1.17% in Polynesia), because of the very low prevalence of leprosy. For the diagnosis of the subclinical infection, our experience based on the follow up of a population of 1,123 contacts, showed that such a control program would have low feasibility and low cost-effectiveness. After 5 years, the proportion of individuals developing the disease was not different among the seropositive than among the seronegative subjects. Finally, only 17.5% of the new cases of leprosy detected, were issued from the contact population followed. In conclusion and in a practical point of view, although this test may contribute to the diagnosis of patients, it is not likely useful neither for the diagnosis nor for the prognosis of leprosy in a population.