Objectives: This research aimed to present a novel glasses-free distance random-dot stereotest system (GFDRDSS) using an eye-tracking method. Methods: A single-view autostereoscopic display applying a backlight control system combined with an eye-tracking method and the corresponding random-dot stereotest software were developed to create a GFDRDSS with a viewing distance of 5 m. The stereoacuity of 12 subjects with normal eye position was evaluated using the Randot Stereotest, Stereoscopic Test Charts vol. 3 (Yan’s Charts), Distance Randot® Stereotest, and GFDRDSS. Results: The GFDRDSS could provide distinct and stable glasses-free stereoscopic perception even while the subject was moving their head. It could evaluate binocular disparities of 40–2,400 arcsec. Eleven subjects with normal near visual acuity had fine near stereovision (20–60 arcsec) using the Randot stereotest and Yan’s Charts. Under refractive correction, 10 subjects had fine stereovision (≤60 arcsec) using the GFDRDSS at a distance of 5 m, and 9 had fine stereovision using the Distance Randot® Stereotest at 3 m. Other subjects described the 100 arcsec-level stereograms correctly. The results exhibited a concordance of stereoacuity within one degrade between the two distance stereotests. Conclusion: The proposed GFDRDSS can alternately project a couple of random-dot stereograms to the subjects’ eyes and provide a glasses-free distance stereotest, which showed good concordance with the Distance Randot® Stereotest. More data are needed for statistical studies.