Lamb mortality is of concern to the New Zealand sheep industry from both an economic and an animal welfare perspective and multiple-born lambs are particularly vulnerable. The relationship between behaviour of twin- (n=850) and triplet-born lambs (n=831) at 12 to 24 hours after birth and survival from tagging until weaning was investigated using data from three different experiments. Lamb behaviours such as the number of high- and low-pitched bleats, time spent walking, standing and sitting, time taken to reach the dam and the contact zone (within 1 metre of the ewe), time spent with its dam and time spent with an alien ewe were monitored in a triangle pen test (Nowak et al., 1987). Data were analysed using logistic regression to examine the effect of lamb behaviour on survival, with twin- and triplet-born lambs analysed separately. Results are presented as the odds ratio with the 95% confidence limit in brackets. For each additional second required for twin-born lambs, but not triplet-born lambs, to reach the contact zone they were 0.989 (0.979-1.008) times as likely to survive (P < 0.05). Whereas, for each additional second required to reach the dam triplet-born lambs, but not twin-born lambs, were 0.994 (0.989-0.999) times as likely to survive (P < 0.05). These results partially support previous research. For every additional high-pitched ‘distress’ bleat emitted twin-born lambs were 0.978 (0.964-0.992) times as likely to survive (P < 0.05), however, this effect was not seen in triplet-born lambs. The time the lambs spent walking, standing, sitting, with its dam or with the alien ewe did not affect survival in twin- or triplet-born lambs (P > 0.05). It was concluded that triplet-born lambs needed to be able to quickly locate its dam, while twin-born lambs only needed to quickly locate the contact zone for their chances of survival to increase.