It is unclear whether blood culture samples should be obtained through one or multiple catheter lumens. We measured how frequently drawing blood cultures from all the lumens from a multilumen catheter resulted in discordant results and how often these caused medical interventions. We performed a retrospective review of the microbiology database of the National Institutes of Health (NIH) Clinical Center. Most patients were immunocompromised. All blood cultures obtained from May 1, 2007 to April 30, 2009 were reviewed. We analyzed all positive blood cultures (i.e., positivity of any of the blood cultures drawn through the catheter lumens) when simultaneous samples from different lumens were obtained, and reviewed the medical charts of those in which blood cultures from different lumens had discordant results (i.e., not all lumens revealed the same organism). We also analyzed how often the discordant results lead to a medical intervention, defined as a change of antimicrobials and/or removal of the catheter. There were 405 episodes of positive blood cultures, in which simultaneous samples of different lumens of a multilumen catheter were obtained. Eighty-five episodes (21 %) were considered to be contaminants and excluded. We analyzed 320 episodes of positive blood cultures in 153 patients; 173 episodes (54.1 %) had discordant results. In 77 % of the 173 episodes, the discordant isolate led to a medical intervention. In immunocompromised patients, sampling all the lumens of a multilumen catheter results in more positive blood cultures, and many of these result in medical interventions. When evaluating bloodstream infection in patients with multilumen catheters, sampling all lumens should be strongly considered.