The aim of this study was to compare two traditional pattern matching techniques, pulsed-field gel electrophoresis (PFGE) and random amplified polymorphic DNA (RAPD), with the more reproducible technique of multilocus sequence typing (MLST) to genotype a blinded sample of Pseudomonas aeruginosa isolates from cystic fibrosis (CF) patients. A blinded sample of 48 well-characterized CF P. aeruginosa isolates was genotyped by PFGE, RAPD, and MLST, each performed in a different laboratory. The discriminatory power and congruence between the methods were compared using the Simpson's index, Rand index, and Wallace coefficient. PFGE and MLST had the greatest congruence with the highest Rand index (0.697). The discriminatory power of PFGE, RAPD, and MLST were comparable, with high Simpson's indices (range 0.973-0.980). MLST identified the most clonal relationships. When clonality was defined as agreement between two or more methods, MLST had the greatest predictive value (100 %) in labeling strains as unique, while PFGE had the greatest predictive value (96 %) in labeling strains as clonal. This study demonstrated the highest level of agreement between PFGE and MLST in genotyping P. aeruginosa isolates from CF patients. MLST had the greatest predictive value in identifying strains as unique and, thus, has the potential to be a cost-efficient, high-throughput, first-pass typing method.