We estimate the incidence rate, serological conversion rate and basic case reproduction number (R0) of Leishmania infantum from a cohort study of 126 domestic dogs exposed to natural infection rates over 2 years on Marajó Island, Pará State, Brazil. The analysis includes new methods for (1) determining the number of seropositives in cross-sectional serological data, (2) identifying seroconversions in longitudinal studies, based on both the number of antibody units and their rate of change through time, (3) estimating incidence and serological pre-patent periods and (4) calculating R0 for a potentially fatal, vector-borne disease under seasonal transmission. Longitudinal and cross-sectional serological (ELISA) analyses gave similar estimates of the proportion of dogs positive. However, longitudinal analysis allowed the calculation of pre-patent periods, and hence the more accurate estimation of incidence: an infection-conversion model fitted by maximum likelihood to serological data yielded seasonally varying per capita incidence rates with a mean of 8.66 x 10(-3)/day (mean time to infection 115 days, 95% C.L. 107-126 days), and a median pre-patent period of 94 (95% C.L. 82-111) days. These results were used in conjunction with theory and dog demographic data to estimate the basic reproduction number, R0, as 5.9 (95% C.L. 4.4-7.4). R0 is a determinant of the scale of the leishmaniasis control problem, and we comment on the options for control.