The results of an intensive survey conducted during 1964-68, which examined the reproductive performance of 184,217 ewes in registered Romney stud flocks throughout New Zealand, are given. The flocks, situated in six geographical localities based on latitude, were not a random sample and represented the upper limits of reproductive performance quoted in an earlier paper. The following are conclusions from the data:-I. There were significant differences between areas in the proportions of 3.5-(P < 0.01), 4.5-(P < 0.01), and 6.5-year and older ewes (P < 0.05).2. The mean incidence of barren ewes was 6.8%, with the following area differences: 1, 9.0%; 2, 5.9%; 3, 7.5%; 4, 6.9%; 5, 5.5%; 6, 5.4% (P < 0.01).3. A mean of 1.8% of ewes died between mating and lambing, and a further 1.0% died during lambing. There were no between-area differences in these losses.4. Highly significant between-area differences were demonstrated in lamb production. These were due to significant decreases from north to south in the proportions of ewes lambing single lambs (P < 0.001), and significant linear increases from north to south in the proportions of ewes lambing multiple births (twins, P < 0.001; triplets, P < 0.01).5. The incidence of ewes which lost lambs and of lamb deaths was lowest in South Island flocks. The following proportions of singles, twins, and triplets were lost through abortion and perinatal death: singles, 12.5%; twins, 8.8%; triplets, 21.0%, with significant differences between areas (P < 0.01 -P < 0.001).6. Analyses of the distribution curves of ewes lambing single and multiple births at 5-day intervals throughout lambing, and of the patterns of live and dead single and twin births, confirmed the large between-area differences.7. There were no significant between-year differences either in the age structure of flocks, the incidence of barrenness, ewe deaths, the proportions of ewes lambing single or multiple births, or of lamb mortality.