Reproductive performance of Dorset ewes in the Cornell University STAR accelerated lambing program was evaluated. This program consists of five 30-d concurrent breeding and lambing periods per year beginning on January 1 (Sl), March 15 (S2), May 27 (S3), August 8 (S4) and October 20 (S5). Optimally, a ewe could lamb every 7.2 mo beginning at 1 yr of age, Records on 1,084 ewes over 7 yr beginning S1, 1982 Reproductive traits considered were ewe fertility, prolificacy, days to first lambing (DFL) and days between lambing (DBL). Fertility and prolificacy differed (P < .001) by exposure seasons. Exposure during favorable seasons (S1, S4, S5) increased fertility (P < .001) and prolificacy (P < .01) over that observed in unfavorable seasons (S2, S3). On average, first lambing occurred at 476 ± 5 d with ewe lambs first exposed during S1, S4 and S5 lambing at younger ages (P < .001). Ewes had more DBL (P < .001) if initial exposure following lambing occurred during an unfavorable season. A curvilinear relationship of ewe age with fertility (P < .001) , prolificacy (P < .01) and DBL (P < .05) was observed in ewes less than 4 yr old. At first exposure following lambing, fertility was lower (P < .001) than at subsequent exposures and ewes that had nursed lambs were less prolific (P < .01) than those that had not. Prolificacy and nursing status had little effect on fertility or DBL (P > .10). Fertility was transformed to an underlying liability scale based on the expected fertility of ewes of a given age and reproductive history. Variance components were estimated by least-squares (LS) and restricted maximum likelihood (REML) procedures. First-exposure fertility was not heritable. heritabilities for fertility at first post-lambing exposure, first-lambing prolificacy, and second-lambing prolificacy were .191 ± .088 (LS) and .200 ± .132 (REML), .914 ± .091 (LS) and .158 ± .144 (REML), and .168 ± .082 (LS) and .210 ± .137 (REML), respectively. Genetic variation in DFL (P < .01) and DBL (P < .001) was found. Although intrasire ewe variation was detected for prolificacy (P < .001) and DBL (P < .01), repeatability estimates were low and smaller than heritability. / Ph. D.