In a previous study, we proposed to associate spoligotyping and typing with the variable number of tandem DNA repeats (VNTR) as an alternative strategy to IS6110-restriction fragment length polymorphism (RFLP) for molecular epidemiological studies on tuberculosis. The aim of the present study was to further evaluate this PCR-based typing strategy and to describe the population structure of Mycobacterium tuberculosis in another insular setting, Sicily. A collection of 106 DNA samples from M. tuberculosis patient isolates was characterized by spoligotyping and VNTR typing. All isolates were independently genotyped by the standard IS6110-RFLP method, and clustering results between the three methods were compared. The totals for the clustered isolates were, respectively, 15, 60, and 82% by IS6110-RFLP, spoligotyping, and VNTR typing. The most frequent spoligotype included type 42 that missed spacers 21 to 24 and spacers 33 to 36 and derived types 33, 213, and 273 that, together represented as much as 26% of all isolates, whereas the Haarlem clade of strains (types 47 and 50, VNTR allele 32333) accounted for 9% of the total strains. The combination of spoligotyping and VNTR typing results reduced the number of clusters to 43% but remained superior to the level of IS6110-RFLP clustering (ca. 15%). All but one IS6110-defined cluster were identified by the combination of spoligotyping and VNTR clustering results, whereas 9 of 15 spoligotyping-defined clusters could be further subdivided by IS6110-RFLP. Reinterpretation of previous IS6110-RFLP results in the light of spoligotyping-VNTR typing results allowed us to detect an additional cluster that was previously missed. Although less discriminative than IS6110-RFLP, our results suggest that the use of the combination of spoligotyping and VNTR typing is a good screening strategy for detecting epidemiological links for the study of tuberculosis epidemiology at the molecular level.