Objectives: Bloodstream fungal infections have increased worldwide during the last two decades and are considered a serious public health problem. A prospective, observational study was conducted at a Portuguese University hospital, aiming to evaluate the susceptibility pattern of isolates from patients with bloodstream fungal infection and to determine the degree of similitude between distinct isolates of the same patient. Methods: Yeasts isolated from blood cultures (89) and from other body sites (40) of patients with fungaemia admitted at a university hospital of Porto were collected and identified with VITEK 2. Minimal Inhibitory Concentrations (MIC) to fluconazole, posaconazole voriconazole, amphotericin B, Caspofungin and Anidulafungin were determined according to the protocol M27-A3 from the Clinical Laboratory for Standards Institute (CLSI). The strains were classified as S, R, susceptible-dose dependent (S-DD) or non susceptible accordingly the CLSI protocol. Sixty strains (C. albicans, C. glabrata and C. parapsilosis) isolated from blood cultures (40) and other biological products (20), from 16 patients were studied regarding the presence of different restriction patterns after restriction endonuclease analysis (REA) with HinfI enzyme. Restriction patterns were analyzed using the UVIDOC 12.6 software and compared among the distinct groups of strains. Results: From a total of 89 strains isolated from blood cultures during the first fungaemia episode 43% corresponded to C. albicans, 26% to C. parapsilosis, 13% to C. glabrata and 7% to C. tropicalis. C. albicans (28%), C. parapsilosis (24%), C. glabrata (18%) and C. krusei (17%) were the most frequent yeasts isolated from other body sites. Regarding the susceptibility profile, 8% of C. albicans, 17% of C. parapsilosis, C. tropicalis and 58% of C. glabrata isolates were resistant to fluconazole. Resistance to equinochandins was detected in 8% of C. glabrata and 22% of C. parapsilosis. Regarding molecular typing, the method did not provide satisfactory results for C.parapsilosis since the same pattern was obtained when comparing among different patients. For the other tested Candida species the results obtained among the different set of isolates for each patient were very heterogeneous. From one patient yielding C. albicans (n=2) and C. parapsilosis n=(9) strains, isolated both from blood cultures and other biological products, the differences both in the restriction pattern and susceptibility profile were only found at a interspecies level. Conversely, the C. albicans strains isolated only from the blood cultures of 2 patients (3 strains of each one), were all different within each patient; in one patient the susceptibility profiles were similar but in the other major differences were registered. Conclusion: High resistance to azoles and equinochandins was observed. Differences in susceptibility pattern do not necessarily imply differences in restriction patterns, i.e. different strains. REA is a rapid and simple technique to be used for strains typing. This technique could be of value in the follow up of patients under antifungal prophylaxis protocols to clarify strains relatedness in the case of emergence of fungal infection. Acknowledgments: S Costa de Oliveira and A P Silva are supported by the grants SFRH/BD/27662/2006 and SFRH/BD/29540/2006, respectively, from Portuguese Science and Technology Foundation (FCT).