Serious infections associated with intravascular catheters are common. The available data suggests there are likely to be more than 500 000 cases of catheter-associated bloodstream infections occurring annually in Western Europe and the USA. These may be associated with as many as 100 000 deaths. The pathophysiology of this common condition is still not fully elucidated. With catheters that are in place for short periods (a few days), microbial migration down the outer surface of the device to the intravascular tip predominates. For catheters that are in place for longer periods, migration occurs more often via the internal lumen. After being in place for more than 8 days, nearly all central vein catheters will have microorganisms embedded in a biofilm within the catheter lumen. In some catheters, microorganisms will proliferate to sufficient numbers for systemic sepsis to result. The occurrence and rate of this proliferation is dependent on microbial virulence factors, host factors, and characteristics of the catheter. Diagnosis of intravascular device-associated sepsis remains problematic because the pathophysiology of the condition changes with time and because standard culture techniques rarely detect organisms embedded in biofilms. The semiquantitative roll method on blood agar remains in common use because of its simplicity. However, the method only samples the external surface of the catheter. For catheters that have been in place for extended periods of time, methods that better sample the internal lumen, such as sonication and quantitative broth methods, should be developed and used.