The influence of the volume of blood cultured on the rate of detection of bacteremia was evaluated in a routine 12-tube blood culture system using 1693 samples from 1502 patients. Blood samples were drawn simultaneously into two transport tubes. The blood volume cultured was the only varying parameter. Generally, 17 % more cultures with clinically significant microorganisms (bothEnterobacteriaceae and gram-positive cocci) were found when blood from two instead of one tube was used (in most cases comparing 13–16 ml of blood with 6.5–8 ml). Of the most prevalent species, the maximum average extra yield was observed forStaphylococcus aureus (26 %) followed byEscherichia coli (16 %) andStreptococcus pneumoniae (12 %). In adults most cases of bacteremia are low-grade. The grade of bacteremia in our patient population was on average as low as 0.25 CFU/ml blood. Therefore, all patients suspected of having bacteremia should have the benefit of a sufficient volume of blood cultured. Since the volume of blood cultured seems to be the single most important factor in the detection of bacteremia, it is imperative that the volume is the same in comparative studies of different blood culture systems.