Forty pregnant Red Maradi goats were used in Caprine Centre of Maradi (Niger) for two subsequent lactations from day one to day 14 of lactation to evaluate the effect of feeding levels of dams on production performances of dams and post-natal performance of kids; to determine the effect of previous levels of feeding dam on production performances of kid and dam; and to determine the ME intakes of kid and dam and their requirements based on AFRC and NRC systems. One diet was formulated for two kiddings as follows: 38% of crushed and urea-treated of millet stover (CUTMS), 32% of groundnut haulms (GH), 22% of wheat bran (WB) and 8% of cottonseed cake (CSC). For the first kidding until the second kidding, four treatments were constituted, T1, T2, T3, and T4, and formed respectively by four feeding levels 842, 934, 1079 and 1300 g/kg DM corresponding to 200 g, 400 g, 600 g and 800 g of milk production, respectively. A randomized block design was applied on 40 does according to body weight, parity and birth type implying 10 does per treatment, and forty-nine born kids of which thirty-one singles and eighteen twins were fed colostrum and milk using feeding bottles. During the second kidding, dams were fed with T4. Dry matter (DM) intake and ME intake of dams increased linearly (P<0.001) with increasing level of feeding. Kids ME intake (kMEint) increased linearly (P<0.05) during the first lactation with increasing level of feeding while there was no variation (P>0.05) of kMEint at the second lactation. The previous level of feeding, age and birth type affected (at least P<0.05) the final body weight and body weight change of dams. The kMEint observed were similar to ME requirements of AFRC but lower than that of NRC. To overcome the ME requirements of kids and lactating does of Red Maradi goats, the requirements of NRC system should be used.