Hospital antibiotic consumption is generally adjusted to occupancy. This study hypothesised that the number of blood culture samples could be a surrogate marker for case-mix adjustment. Antibiotic consumption was compared over 16 consecutive trimesters in one medical ward in terms of patient-days or blood culture samples. Compared with patient-days, measurement adjusted to blood culture samples detected three trimesters with an unusually high consumption, and one trimester with consumption falsely classified as high because of a high incidence of infections. Blood culture numbers enabled easy and accurate identification of periods with a drift in antibiotic consumption ina medical ward.