While the gold standard for molecular epidemiological studies on tuberculosis is changing towards MIRU-VNTR typing because this technique generates easily analyzed numerical results, it is less labor intensive and has a discriminative power comparable to that of IS6110-based RFLP, especially when 24 loci are analyzed; more extensive and representative validation studies are needed to confirm this. In this study we genotyped 41 Mycobacterium tuberculosis isolates, about 40% of the total case load of the study year, from the Warao people, an indigenous population who live in a geographically isolated area in Venezuela and have a high TB incidence of 450/100,000. IS6110-based RFLP analysis on these isolates indicates that 78% of the strains are in clusters, suggesting a very high transmission rate. We show that both the 15-locus MIRU-VNTR combined with spoligotyping, as well as the 24-locus MIRU-VNTR typing have sufficient discrimination power (an HGI of 0.93 and 0.95, respectively) to replace IS6110-based RFLP (HGI=0.93) and thus are useful tools to study the molecular epidemiology of tuberculosis in this high TB incidence population.