The aims of this cross-sectional prevalence study were an epidemiologic survey of bacterial flora in older residents of long-term care facilities (LTCFs) in the Attika Region; an attempt to define the factors leading to colonization of urine, nasopharynx and decubitus ulcers of residents with various microorganisms; and, finally, obtaining data on the resistance pattern of these. During the two year period June 2002-May 2004, samples of urine, anterior nares and decubitus ulcers were collected from 1523 residents in 41 LTCFs. There were 863 gram-negative isolates, the leading organisms being Escherichia coli, Proteus mirabilis and Klebsiella pneumoniae. Also, there were 244 gram-positive isolates, the most frequently seen of which were Staphylococcus aureus. Commensals of urine and decubitus ulcers were predominantly Gram-negative bacteria and nares Staphylococcus aureus. 841 (55.2%) residents were taking, or had within the previous 3 months been taking, a systemic antibiotic. Quinolones were the most frequently prescribed antibiotic class (36%). The institutional prevalence of urinary catheterization averaged 21.4%. From the total studied population, 27.6% and 12% were carriers of bacteria resistant to broad spectrum b-lactams and fluoroquinolones respectively. Only 3.4% were carrying MRSA in their nares. Urinary catheterization, nasogastric feeding tubes, decubitus ulcers, lack of mobility and history of type 2 diabetes mellitus were the most important factors favoring colonization. Logistic regression analysis revealed that the single factor affecting the pattern of resistance of the isolated microorganisms was the type of antimicrobial administered the last trimester. Bacterial flora of LTCF residents differed according to the level of hygiene of the LTCF and the site of colonization. Significant antibiotic pressure exists in long-term care facilities, a fact that is reflected in antibiotic resistance patterns. Factors like urinary catheterization, the presence of feeding tubes and underlying diseases impose significantly on the process of colonization. The results presented may have implications for empirical antimicrobial prescribing for patients in LTCFs with poor perceived hygiene standards, who have urinary catheters in situ and a history of recent antimicrobial intake. Broad spectrum antimicrobials may be more appropriate for these patients if indicated.