Contact surveillance is a valuable strategy for controlling leprosy. A dynamic cohort study of leprosy contacts was initiated in 1987 at Oswaldo Cruz Foundation. The objective of this work was to review the data on the major risk factors leading up to the infectious stage of the disease, estimate incidence rates of leprosy in the cohort and characterise the risk factors for the disease among the contacts under surveillance. The incidence rate of leprosy among contacts of leprosy patients was estimated at 0.01694 cases per person-year in the first five years of follow-up. The following factors were associated with acquiring the disease: (i) not receiving the BCG vaccine, (ii) a negative Mitsuda reaction and (iii) contact with a patient with a multibacillary clinical form of leprosy. The contacts of index patients who had high bacilloscopic index scores > 1 were at especially high risk of infection. The following factors were associated with infection, which was defined as a seropositive reaction for anti-phenolic glicolipid-1 IgM: (i) young age (< 20 years), (ii) a low measured Mitsuda reaction (< 5 mm) and (iii) contact with an index patient who had a high bacilloscopic index. BCG vaccination and re-vaccination were shown to be protective among household contacts. The main conclusions of this study indicate an urgent need for additional leprosy control strategies in areas with a high incidence of the disease.