To assess the long-term safety and effectiveness of non penetrating deep sclerectomy (DS) and to compare the incidence and the severity of postoperative complications and the IOP results according to surgical adjuvants (implant device, antimetabolite or both) were used or not.Retrospective non randomised study including 171 eyes (136 patients), mean age: 63.9 years) with medically uncontrolled open-angle glaucoma and without previous filtering surgery. 81 eyes (48.2%) had severe glaucomatous damage. All procedures were performed according to the Kozlov's and Memoud's technique. Except for 8 eyes, they were associated with the placement of an implant device (SKGEL or T-FUX) and/or intra-operative application of low dose antimetabolite (5-FU in 58 eyes and mitomycine C in 53 eyes).Mean follow-up was 39.6 +/- 18.3 months. According to surgery, DS were categorized in 4 groups: Group 1: DS with Healon GV (n=8)(4.7%); Group 2: DS with antimetabolite application (n=26) (15.2%); Group 3: DS with placement of an implant (n=53) (31%). Group 4: intraoperative antimetabolite +implant device (n=84 eyes) (49.1%). Peroperative microperforations without iris hernia occurred in 35 eyes (21%). 1st month postoperative complications were observed in 90 eyes (52.6%) with mild to moderate hyperfiltration in 27 eyes, excessive scarring of filtration bleb in 38 eyes, and iris incarceration in 10 eyes. 5-FU injections were given in 58 eyes (34%). YAG gonioperforation was needed in 107/171 eyes (63%) and was complicated by iris incarceration in 9 eyes. Early and late spontaneous iris incarceration was observed in 10 eyes. A second filtering procedure was needed in 10 eyes.