Montenegro skin tests were carried out in 1985 and 1987 on two groups of school children in the city of Rio de Janeiro. Group A consisted of 449 children residing in the Jacarepagua district, in areas where transmission of human and canine American cutaneous leishmaniasis (ACL) has been high; this group was considered to be the one at greatest risk of acquiring the infection through L. braziliensis. Group B consisted of 282 children from Bonsucesso, a suburb of Rio de Janeiro that is at a considerable distance from any area where ACL is endemic; this was a lowest risk group, and was thus used as a control. Analysis of the cutaneous test results showed that in Group A, 8.9% of the sample tested positive, whereas in Group B the result was only 2.1%. In group A there were no significant differences in the proportion of positives when analysed according to sex, age and the year when the tests were carried out. A study of the effects associated with place of residence together with other variables was carried out using log-linear regression analysis. It showed that effects arising from place of residence were maintained; that this was the only significant effect; and that it was independent of other variables. The testing of school children using the Montenegro intradermal test was shown to be a useful procedure in the characterization of localities in which there is a risk of ACL infection.