To assess the feasibility of evisceration with primary implant placement in patients with endophthalmitis. Pilot retrospective study. Sixty-seven patients with endophthalmitis. A retrospective study was conducted to review the files of 67 patients with endophthalmitis who underwent evisceration with primary implant placement over an 8-year period. These patients' files were reviewed to evaluate the following: persistent infection and inflammation, spread of infection to contiguous or remote sites, implant exposure or extrusion, and successful fitting of the prosthesis. Quiescence of the infection, complete healing of the socket, and long-term retention of the implant. Sixty-three patients successfully retained their primary implant with rapid resolution of infection and inflammation. In 1 patient, attempts at implant placement were abandoned during surgery because of marked scleral necrosis. Delayed implant extrusion was noted in 2 patients, 10 and 12 days after surgery, respectively. In 1 diabetic patient, an orbital abscess developed that was evacuated with implant exchange. In a fourth patient, marked conjunctival prolapse developed in the early postoperative period with eventual obliteration of the inferior fornix. The patient declined further management. Minor complications included a central conjunctival dehiscence or necrosis (n = 2) and a pyogenic granuloma that was excised (n = 1). All minor complications healed without sequelae. Overall, 12 % of patients experienced complications (n = 8). Based on the low overall complication rate in the present study (12%), the ease of evisceration, the convenience of a single procedure, the rapid resolution of infection, and the successful retention of the implant in most of our patients, it is plausible to offer evisceration with primary implant placement to patients with recalcitrant endophthalmitis or panophthalmitis. This would avoid a secondary surgical intervention in most patients, as opposed to delayed secondary implantation, in which case 100% of the patients would require, at least theoretically, another intervention.