Leptospirosis is a bacterial zoonosis with world distribution and multiform clinical spectrum in men and animals. The etiology of this disease is the pathogenic species of Leptospira, which cause diverse manifestations of the disease, from mild to serious, such as the Weil disease and the lung hemorrhagic syndrome with lethal proportions of 10% - 50%. This is an emerging problem of urban health due to the growth of marginal neighborhoods without basic sanitary conditions and an increased number of rodents. The presence of rodents and the probability of having contact with their urine determine the likelihood for humans to get infected. In this paper, we simulate the spatial distribution of risk infection of human leptospirosis according to the proximity to rodent burrows considered as potential source of infection. The Bessel function K₀ with an r distance from the potential point source, and the scale parameter α in meters was used. Simulation inputs were published data of leptospirosis incidence rate (range of 5 to 79 x 10 000), and a distance of 100 to 5000 meters from the source of infection. We obtained an adequate adjustment between the function and the simulated data. The risk of infection increases with the proximity of the potential source. This estimation can become a guide to propose effective measures of control and prevention.