OBJECTIVE AND METHOD TO compare the utility of an ELISA using 3 recombinant antigens with that of the skin biopsy to estimate incidence of infections in a sentinel cohort of individuals living in an endemic community in southern Mexico during a set of 11 subsequent ivermectin treatments. RESULTS The apparent community prevalence of infection and microfilarial skin infection before and after 11 treatments with ivermectin plus nodulectomy were 78% and 13%, and 0.68 mf/mg and 0.04 mf/mg, respectively, as measured by skin biopsy. Of a group of 286 individuals participating in all surveys, a sentinel cohort of 42 mf and serologically negative individuals had been followed since 1994. The annual percentage of individuals becoming positive in this cohort was 24% (10/42), 28% (9/33), 0%, and 4.3% (1/23) in 1995, 1996, 1997 and 1998, respectively. Likewise, the incidence in children 5 years and under (n = 13) within this sentinel cohort was 15% (2/13), 18% (2/11), 0% and 11% (1/9), respectively. All individuals became positive to both tests simultaneously, indicating that seroconversion assessed infection incidence as accurately as skin biopsy in the sentinel group. CONCLUSION Incidence monitoring of a sentinel cohort provides an estimation of the parasite transmission in the community; it is less costly than massive sampling, and a finger prick blood test might be more acceptable in some communities.