This study aimed to determine the effects of different lengths of progesterone and equine chorionic gonadotropin (eCG) treatment on reproductive performance during the non-breeding season in ewes. Progesterone-impregnated sponges were inserted intravaginally for 7 d in Group 1 (G1, n = 24), 9 d in Group 2 (G2, n = 25), and 11 d in Group 3 (G3, n = 24). On the day of sponge removal, eCG (350 IU) was injected and 24 h later the ewes were exposed to rams. Pregnancy was diagnosed by transrectal ultrasonography 30 ± 3 d after mating. Oestrus symptoms of G2 and G3 started earlier than G1, but there was only a significant difference between G1 and G3 (P = 0.013). In G1, G2, and G3, the oestrus responses were 100%, 100%, and 95.8%, pregnancy rates were 79.2%, 72%, and 75%, lambing rates were 79.8%, 72%, and 75%, litter sizes were 2.5, 2.2, and 2.4, and multiple birth rates were 86.7%, 80%, and 85.7%, respectively (P > 0.05). In conclusion, the short-term progesterone-impregnated sponge and eCG treatment provided excellent oestrus induction. The effects of different days (7, 9, or 11) of progesterone treatment on reproductive indicators were similar; therefore, the stress caused by the inserted sponge can be reduced by short-term use.