The purpose of the study was to compare the IS6110-RFLP (RFLP) results obtained in a previous epidemiological study in the city of Barcelona, Spain [Borrell S, Espanol M, Orcau A, Tudo G, March F, Cayla JA, et al. Factors associated with differences between conventional contact tracing and molecular epidemiology in study of tuberculosis transmission and analysis in the city of Barcelona, Spain. J Clin Microbiol 2009 Jan;47(1):198-204.] with the results obtained with IS6110-fAFLP, [Thorne N, Evans JT, Smith EG, Hawkey PM, Gharbia S, Arnold C. An IS6110-targeting fluorescent amplified fragment length polymorphism alternative to IS6110-restriction fragment length polymorphism analysis for Mycobacterium tuberculosis DNA fingerprinting. Clin Microbiol Infect 2007 Oct;13(10):964-70.] on the ability to detect recent transmission. fAFLP was applied to DNA samples of RFLP clustered strains, with and without known epidemiological links, with the additional inclusion of four nucleotide-specific fluorophores to further increase the discrimination of the fragments obtained. Four-colour fAFLP was performed on 123 RFLP clustered strains with no epidemiological link (NELC) and on 28 epidemiologically linked RFLP clustered (ELC) strains grouped into 48 and 13 clusters respectively. Clustering results obtained by the two methods were highly congruent in ELC strains with fAFLP allocating 92.3% of the ELCs. For the NELCs, RFLP results were confirmed in 39/48 (81.2%) of fAFLP-clusters with 0-1 different fragments and 9/48 (18.8%) differed in 2-4 fragments, which are considered genetically related but not recently transmitted. In conclusion, overestimation of recent tuberculosis transmission can occur because of the inaccurate analysis of RFLP results. Four-colour fAFLP allows us to differentiate between recent transmission strains and epidemiologically unrelated but genetically related strains.