There is currently a lack of prospective studies comparing multiple treatment options for single-sided deafness (SSD) in terms of long-term sound localization outcomes. This randomized controlled trial (RCT) aims to compare the objective and subjective sound localization abilities of SSD patients treated with a cochlear implant (CI), a bone conduction device (BCD), a contralateral routing of signals (CROS) hearing aid, or no treatment after two years of follow-up. About 120 eligible patients were randomized to cochlear implantation or to a trial period with first a BCD on a headband, then a CROS (or vice versa). After the trial periods, participants opted for a surgically implanted BCD, a CROS, or no treatment. Sound localization accuracy (in three configurations, calculated as percentage correct and root-mean squared error in degrees) and subjective spatial hearing (subscale of the Speech, Spatial and Qualities of hearing (SSQ) questionnaire) were assessed at baseline and after 24 months of follow-up. At the start of follow-up, 28 participants were implanted with a CI, 25 with a BCD, 34 chose a CROS, and 26 opted for no treatment. Participants in the CI group showed better sound localization accuracy and subjective spatial hearing compared to participants in the BCD, CROS, and no-treatment groups at 24 months. Participants in the CI and CROS groups showed improved subjective spatial hearing at 24 months compared to baseline. To conclude, CI outperformed the BCD, CROS, and no-treatment groups in terms of sound localization accuracy and subjective spatial hearing in SSD patients. Netherlands Trial Register (https://onderzoekmetmensen.nl): NL4457, trial.