Radical prostatectomy has in randomised settings failed to demonstrate a survival difference in patients with low-risk prostate cancer when tested against an observational strategy. Active surveillance has been introduced in order to reduce overtreatment by distinguishing between cancers with a biological potential, and truly indolent cancers best left untreated. Preliminary results from large prospective active surveillance cohorts are promising; however, uncertainties persist concerning optimal patient selection and follow-up, as well as the long-term safety.