Two experiments were carried out to determine the effect of slope on lamb mortality. In Experiment 1, 166 Marshall Romney (MR), 170 conventional Romney (CR), and 59 MR X CR ewes were allocated to one of 17 lambing areas of 0.2-0.4 ha and uniform slope. Mean slopes ranged from 3 to 44 ° and each area was surrounded by electric netting fences. Ewes entered their respective lambing areas 2 days before the expected lambing date and remained there for 4-5 days. In Experiment 2, two 1 ha plots selected on the basis of conformity of slope (36--44°) and 100-120 m long were each divided into three equal lambing areas by fences down the slope. Barriers either 15 or 30 m apart were constructed of 33-cm-high woven polyethylene mesh across the slope on two of the three lambing areas in both plots. Lamb mortality in Experiment 1 was not affected by slopes up to 30° but there was a large increase in lamb mortality with slopes of 36--44° (19.4% lamb mortality for less than 30° versus 50.1 % lamb mortality for more than 30°). In Experiment 2, 83-90% of lambs slipped off the birth site and 68-70% of ewes followed their lambs down the slope. The barriers stopped only 60% of the lambs encountering them. The other 40% passed under or over the barriers. Barriers increased separation of twin lambs (39% no barriers versus 65% barriers). The incidence of ewes rearing both twins when the co-twins were separated soon after birth by 0-1 m, 2-4 m ,or greater than 4 m was 15/16, 1/5, and 0/18, respectively. MR ewes had a lower lamb mortality than CR ewes (17 versus 43%). Although there were no differences in the incidence oflambs slipping off the birth site (83% MR versus 89% CR), the MR ewes had a lower incidence of abandonment of lambs (20% MR versus 41% CR).