Differentiation between lethal and non-lethal prostate cancer subtypes has become a very important issue in avoiding excessive treatment in an era when prostate-specific antigen (PSA) screening has reduced the rate of prostate cancer deaths by more than 20%. However, it is difficult to determine the patients who may or may not benefit from immediate treatment interventions at the time of the initial diagnosis. The selection of candidate patients who can postpone immediate treatment and undergo follow-ups with a specific surveillance program, or 'active surveillance,' is a practical way to minimize overtreatment. In this review, the benefits and risks of active surveillance are discussed. Future perspectives, including imaging and new biomarkers for improving the outcomes of active surveillance programs, are also discussed.