Dormer (n = 166) and SA Mutton Merino (n = 147) ewes were observed continuously during lambing in the autumn lambing seasons of 1989-1991. Although SA Mutton Merino ewes had larger (P ≤0.01) litters at birth than Dormers (2.08 vs. 1.75), higher (P ≤0.05) levels of mortality at birth (0.101 vs. 0.031) and from 3 days to weaning (0.180 vs. 0.110) resulted in this difference being eliminated at weaning (1.35 vs. 1.37). SA Mutton Merino ewes generally took longer to give birth (P ≤ 0.01; 92 vs. 67 min) and were more likely to be assisted at birth (P ≤0.01; 0.231 vs. 0.102) than Dormers. Post-mortem results suggested that birth stress was more prevalent (P = 0.053) amongst 61 perinatal deaths in SA Mutton Merino lambs than in 30 Dormers (0.575 vs. 0.333). Mortalities within ca. 1 week of birth amongst assisted lambs (pooled across breeds) were generally higher than in their unassisted contemporaries, significantly (P ≤0.05) so in singles (0.444 vs. 0.044) and triplets (0.604 vs. 0.156). Length of parturition was found to be repeatable (t = 0.36 ± 0.07). Permanent and temporary separation of ewes from ≥ 1 lamb were independent of breed, but were lower (P ≤0.05) amongst ewes caring for singles (0.011) than in ewes caring for twins (0.197) or triplets (0.323). Permanent separation was caused by interference and desertion in respectively 0.500 and 0.239 of affected ewes, while lambs followed a non-interfering ewe in 0.174 of cases. Separation was attributed to udder malfunction in 0.043 of the ewes. The incidence of permanent separation caused by interference by other ewes was higher (P ≤ 0.05) in Dormer ewes than in SA Mutton Merinos (0.654 vs. 0.300). Ewes moving from their birth sites ≤120 min after birth were more likely (P ≤0.01) to be permanently separated from ≥1 lamb than ewes which remained longer (0.400 vs. 0.134). This observation was mainly caused by ewes classified as deserting their lamb(s). These ewes left their birth sites sooner (P ≤ 0.01) than those that were not separated from any lambs (156 vs. 351 min). Separation of ewes from ≥1 lamb was related to the density of lambed ewes (r = 0.56) increasing by 2.01% (SEb = 0.35) for 1 lambed ewe/ha. Udder malfunction was recorded in 18 ewes, and was accompanied by high levels of lamb mortality (0.643). Selection against rearing failure should be investigated for the improvement of rearing ability and lamb survival in these flocks, before further attempts are made to increase fecundity. Management should facilitate for ewes remaining on their birth sites for a long time after birth, and avoid stocking densities > 11 pregnant ewes/ha at lambing.