In severe shoulder dystocia,when in itial maneuvers fail,either episiotomy or fetal manipula tion(Rubin,Woods’screw,or posterior arm release)is recommended.We sought to compare maternal and neonatal outcomes between severe shoulder dystocia deliverie s managed with episioto-my versus fetal manipulation.We identified severe shoul-der dystocia deliveries from three d atabases:all shoulder dystocia deliveries(1993-2003and 1994-1997)from two teaching institutions and litig ated cases of shoulder dystocia -associated permanent bra chial plexus palsy from multiple U.S.institutions.Pair -wise comparisonswere made among three groups of deliverie s:those managed by fetal manipulation without episiotomy (fetal manipula-tion -only),those managed by episiotomy withou t fetal manipulation(episiotomy -only),and those managed with both(episiotomy +fetal manipulation).Rates of brachial plexus palsy,neonatal dep ression,and anal sphincter trauma were compared amon g groups usingχ2,with significance at P<.05.Among episiotomy -only,13of 22(59.1%)sustained brachial plexus palsy,co mpared with 20of 57(35.1%)among fetal manipulation -only(P =.05).Twenty -eight of 48(58.3%)-in epi-siotomy +fetal manipulation had bra chial plexus palsy,which did not differ from episiotomy -only(P =.95)but was higher than fetal manipulation -only(P =.02),suggesting that the addition of epis iotomy conferred no benefit in averting neonatal injury.Anal sphincter trauma was significantly more common among episiotomy -only and episiotomy +fetal manipulation,compared with fetal manipulation -only.In severe shoul der dystocia,if fetal manipulation can be performedwithout episiotomy,severe perineal trauma can be averted without incurring greater risk of brachial plexus palsy. In severe shoulder dystocia, when in itial maneuvers fail, either episiotomy or fetal manipula tion (Rubin, Woods’ crew, or posterior arm release) is recommended. We sought to compare maternal and neonatal outcomes between severe shoulder dystocia deliverie s managed with episioto- my versus fetal manipulation. Identified severe shoul-der dystocia deliveries from three d atabases: all shoulder dystocia deliveries (1993-2003 and 1994-1997) from two teaching institutions and litig ated cases of shoulder dystocia -associated permanent bra chial plexus palsy from multiple US institutions. Pair-wise comparisonswere made among three groups of deliverie s: those managed by fetal manipulation without episiotomy (fetal manipula tion-only), those managed by episiotomy withou t fetal manipulation (episiotomy-only), and those managed with both (episiotomy + fetal manipulation). Rates of brachial plexus palsy, neonatal dep ression, and anal sphincter trauma were compared amon g groups usingχ2, with significance at P <.05.Among episiotomy-only, 13 of 22 (59.1%) sustained brachial plexus palsy, co mpared with 20 of 57 (35.1%) among fetal manipulation-all (P = ) -in epi-siotomy + fetal manipulation had bra chial plexus palsy, which did not differ from episiotomy-only (P = .95) but was higher than fetal manipulation-only (P = .02), suggesting that the addition of epis iotomy conferred no benefit in averting neonatal injury. Anal sphincter trauma was significantly more common among episiotomy-only and episiotomy + fetal manipulation, compared with fetal manipulation-only. severe abdominal dermatosis, if fetal manipulation can be performed with episiotomy, severe perineal trauma can be averted without incurring greater risk of brachial plexus palsy.