An evaluation of the utility of rep PCR typing compared to the 15 loci discriminatory set of MIRU-VNTR was undertaken. Twenty-nine isolates of Mycobacterium tuberculosis from patients were examined. Genomic DNA was extracted from the isolates by standard method. The number of copies of tandem repeats of the 15 MIRU-VNTR loci was determined by PCR amplification and agarose gel electrophoresis of the amplicons. M. tuberculosis outbreak-related strains were distinguished from other isolates. MIRU-VNTR typing identified 4 major clusters of strains. The same isolates clustered together after RFLP typing, but rep-PCR identified only 3 of them. The concordance between RFLP and MIRU-VNTR typing was complete, with the exception of two isolates with identical RFLP patterns that differed in the number of tandem repeat copies at two MIRU-VNTR alleles. A further isolate, even sharing the same RFLP pattern, differed by one repeat from the rest of its cluster. We also tested the use of an automated rep-PCR for clinical laboratory applications but it failed to identify the link between two pairs of epidemiologically related strains clustered by the other 2 techniques. For superior discrimination, ease of comparison of results and lower cost, MIRU-VNTR typing should be the favored PCR-based typing tool.