Between September 1987 and October 1990, 265 eyes with rhegmatogenous retinal detachment and severe (with a classification of at least C-3) proliferative vitreoretinopathy were treated with vitrectomy and randomized to treatment with perfluoropropane gas or silicone oil; 131 eyes had undergone no prior vitrectomy (group 1) while 134 eyes had undergone vitrectomy with intraocular gas tamponade (group 2). At the last examination, there were no differences between perfluoropropane gas and silicone oil in achieving visual acuity greater than or equal to 5/200 (43% vs 45% for group 1, 38% vs 33% for group 2) and complete posterior retinal reattachment (73% vs 64% for group 1, 73% vs 61% for group 2). For group 1 eyes followed up for at least 18 months, there was an advantage favoring perfluoropropane gas in achieving complete posterior retinal reattachment (83% vs 60% at 36 months, P = .045). The rates of reoperation and keratopathy were similar, while hypotony was more prevalent in eyes randomized to perfluoropropane gas (group 2). Regardless of tamponade, groups 1 and 2 had similar anatomic and visual success. However, hypotony was twice as prevalent in group 2 (perfluoropropane), and the prevalence of keratopathy increased with follow-up in group 2 (either tamponade). Either tamponade produced better results than those seen with sulfur hexafluoride gas (Silicone Study Report 1).