PURPOSE: To compare pneumatic retinopexy and scleral buckling for repair of primary rhegmatogenous retinal detachment with respect: to visual outcome, single-procedure reattachment rate, and development of proliferative vitreoretinopathy METHODS: A consecutive series of eyes initially treated with pneumatic retinopexy (n = 56) between March 1986 and February 1996 were compared with a selected group of eyes treated with scleral buckling (n = 86) with similar location and distribution of retinal breaks and absence of proliferative vitreoretinopathy, A regression model was developed to adjust for underlying differences between treatment groups, resulting in a cohort of 50 eyes in each group for final comparison. A minimum follow-up of 6 months was obtained. RESULTS: Single-procedure reattachment rare was significantly higher for scleral buckle eyes (42 of 50 eyes, 84%) than for pneumatic retinopexy eyes (31 of 50 eyes, 62%; P less than or equal to.01), Correspondingly, reoperation rate was significantly higher for pneumatic retinopexy eyes (19 of 50 eyes, 38%) than for scleral buckle eyes (7 of 50 eyes, 14%; P less than or equal to.01). Multiple regression analysis evaluating perioperative factors demonstrated that the use of pneumatic retinopexy was the sole factor predictive of retinal detachment after a single procedure (relative odds = 2.20, P =.02). Final reattachment rate, after reoperations, was 98% (49 of 50 eyes) in each group. Except for nonphakic eyes, final visual outcome and rate of postoperative proliferative virreoretinopathy development did not differ significantly between the two procedures. CONCLUSIONS: In phakic eyes, pneumatic retinopexy was associated with a significantly higher reoperation rate than scleral buckling, but resulted in equivalent final visual outcome and reattachment rare after reoperations. If used, it must be incorporated into a strategy in which patient and physician are prepared for a greater chance of reoperation compared to initial management with scleral buckling. (Am J Ophthalmol 1998;126: 658-668. (C) 1998 by Elsevier Science Inc, All rights reserved.).