The effects of melatonin implants on out-of-season breeding in New Zealand Romney composite ewes, was determined by comparison of reproductive performance in ewes treated with progesterone+equine chorionic gonadotrophin (eCG) (control; n=107), melatonin+progesterone+eCG (n=97) or melatonin+progesterone (n=96). Conception rates in melatonin+progesterone+eCG-treated ewes (67%) were higher than in the control ewes (P<0.01; 47%). Pregnancy rates were higher in melatonin+progesterone+eCG-treated ewes (55%; P<0.001) compared with the control ewes (40%). Fewer melatonin+progesterone-treated ewes displayed oestrus (14%; P<0.001) and subsequently became pregnant (6%). Oestrus rates in melatonin+progesterone-treated ewes (14%) were lower than both the melatonin+progesterone+eCG-treated (82%) and control ewes (86%; P<0.001), which were similar to each other. The number of foetuses per pregnant ewe was similar in all three treatment groups. Serum melatonin concentrations at Day -9 were higher in the ewes treated with melatonin and there was a large variation between individual ewes, but concentrations were similar for pregnant and nonpregnant ewes. The combination of higher conception rate and the trend for more lambs per pregnant ewes resulted in more lambs being born per ewe treated in melatonin+progesterone+eCG-treated ewes compared to the other two treatment groups. These results suggest that melatonin implants, in conjunction with administration of progesterone and eCG, may be suitable as a means of increasing the number of lambs born per ewe treated in an out-of-season breeding program in New Zealand sheep flocks while melatonin and progesterone is not.