The number needed to treat (NNT) is a useful way for clinicans to describe the benefit or harm of a treatment as well as the costs involved. When interpreting the NNT it is essential to use the NNT in a clinically equivalent and appropriate setting. When evaluating the consequences of a treatment clinicians should make sure that the patients being treated have the same risk profile asthe study patients. Differences in duration of follow-up and baseline risks can cause significant changes in the NNT; therefore, NNT should be evaluated in addition to relative risk differences and baseline risk to reduce any ambivalence in the assessment of a treatment. This review provides insights into the assessment and clinical use of NNT in the practice.